--- /dev/null
+import functools
+
+from .common import InfoExtractor
+from ..networking import HEADRequest
+from ..utils import (
+ float_or_none,
+ int_or_none,
+ join_nonempty,
+ parse_qs,
+ update_url_query,
+)
+from ..utils.traversal import traverse_obj
+
+
+class RedCDNLivxIE(InfoExtractor):
+ _VALID_URL = r'https?://[^.]+\.(?:dcs\.redcdn|atmcdn)\.pl/(?:live(?:dash|hls|ss)|nvr)/o2/(?P<tenant>[^/?#]+)/(?P<id>[^?#]+)\.livx'
+ IE_NAME = 'redcdnlivx'
+
+ _TESTS = [{
+ 'url': 'https://r.dcs.redcdn.pl/livedash/o2/senat/ENC02/channel.livx?indexMode=true&startTime=638272860000&stopTime=638292544000',
+ 'info_dict': {
+ 'id': 'ENC02-638272860000-638292544000',
+ 'ext': 'mp4',
+ 'title': 'ENC02',
+ 'duration': 19683.982,
+ 'live_status': 'was_live',
+ },
+ }, {
+ 'url': 'https://r.dcs.redcdn.pl/livedash/o2/sejm/ENC18/live.livx?indexMode=true&startTime=722333096000&stopTime=722335562000',
+ 'info_dict': {
+ 'id': 'ENC18-722333096000-722335562000',
+ 'ext': 'mp4',
+ 'title': 'ENC18',
+ 'duration': 2463.995,
+ 'live_status': 'was_live',
+ },
+ }, {
+ 'url': 'https://r.dcs.redcdn.pl/livehls/o2/sportevolution/live/triathlon2018/warsaw.livx/playlist.m3u8?startTime=550305000000&stopTime=550327620000',
+ 'info_dict': {
+ 'id': 'triathlon2018-warsaw-550305000000-550327620000',
+ 'ext': 'mp4',
+ 'title': 'triathlon2018/warsaw',
+ 'duration': 22619.98,
+ 'live_status': 'was_live',
+ },
+ }, {
+ 'url': 'https://n-25-12.dcs.redcdn.pl/nvr/o2/sejm/Migacz-ENC01/1.livx?startTime=722347200000&stopTime=722367345000',
+ 'only_matching': True,
+ }, {
+ 'url': 'https://redir.atmcdn.pl/nvr/o2/sejm/ENC08/1.livx?startTime=503831270000&stopTime=503840040000',
+ 'only_matching': True,
+ }]
+
+ """
+ Known methods (first in url path):
+ - `livedash` - DASH MPD
+ - `livehls` - HTTP Live Streaming
+ - `livess` - IIS Smooth Streaming
+ - `nvr` - CCTV mode, directly returns a file, typically flv, avc1, aac
+ - `sc` - shoutcast/icecast (audio streams, like radio)
+ """
+
+ def _real_extract(self, url):
+ tenant, path = self._match_valid_url(url).group('tenant', 'id')
+ qs = parse_qs(url)
+ start_time = traverse_obj(qs, ('startTime', 0, {int_or_none}))
+ stop_time = traverse_obj(qs, ('stopTime', 0, {int_or_none}))
+
+ def livx_mode(mode):
+ suffix = ''
+ if mode == 'livess':
+ suffix = '/manifest'
+ elif mode == 'livehls':
+ suffix = '/playlist.m3u8'
+ file_qs = {}
+ if start_time:
+ file_qs['startTime'] = start_time
+ if stop_time:
+ file_qs['stopTime'] = stop_time
+ if mode == 'nvr':
+ file_qs['nolimit'] = 1
+ elif mode != 'sc':
+ file_qs['indexMode'] = 'true'
+ return update_url_query(f'https://r.dcs.redcdn.pl/{mode}/o2/{tenant}/{path}.livx{suffix}', file_qs)
+
+ # no id or title for a transmission. making ones up.
+ title = path \
+ .replace('/live', '').replace('live/', '') \
+ .replace('/channel', '').replace('channel/', '') \
+ .strip('/')
+ video_id = join_nonempty(title.replace('/', '-'), start_time, stop_time)
+
+ formats = []
+ # downloading the manifest separately here instead of _extract_ism_formats to also get some stream metadata
+ ism_res = self._download_xml_handle(
+ livx_mode('livess'), video_id,
+ note='Downloading ISM manifest',
+ errnote='Failed to download ISM manifest',
+ fatal=False)
+ ism_doc = None
+ if ism_res is not False:
+ ism_doc, ism_urlh = ism_res
+ formats, _ = self._parse_ism_formats_and_subtitles(ism_doc, ism_urlh.url, 'ss')
+
+ nvr_urlh = self._request_webpage(
+ HEADRequest(livx_mode('nvr')), video_id, 'Follow flv file redirect', fatal=False,
+ expected_status=lambda _: True)
+ if nvr_urlh and nvr_urlh.status == 200:
+ formats.append({
+ 'url': nvr_urlh.url,
+ 'ext': 'flv',
+ 'format_id': 'direct-0',
+ 'preference': -1, # might be slow
+ })
+ formats.extend(self._extract_mpd_formats(livx_mode('livedash'), video_id, mpd_id='dash', fatal=False))
+ formats.extend(self._extract_m3u8_formats(
+ livx_mode('livehls'), video_id, m3u8_id='hls', ext='mp4', fatal=False))
+
+ time_scale = traverse_obj(ism_doc, ('@TimeScale', {int_or_none})) or 10000000
+ duration = traverse_obj(
+ ism_doc, ('@Duration', {functools.partial(float_or_none, scale=time_scale)})) or None
+
+ live_status = None
+ if traverse_obj(ism_doc, '@IsLive') == 'TRUE':
+ live_status = 'is_live'
+ elif duration:
+ live_status = 'was_live'
+
+ return {
+ 'id': video_id,
+ 'title': title,
+ 'formats': formats,
+ 'duration': duration,
+ 'live_status': live_status,
+ }
--- /dev/null
+import datetime
+
+from .common import InfoExtractor
+from .redge import RedCDNLivxIE
+from ..utils import (
+ clean_html,
+ join_nonempty,
+ js_to_json,
+ strip_or_none,
+ update_url_query,
+)
+from ..utils.traversal import traverse_obj
+
+
+def is_dst(date):
+ last_march = datetime.datetime(date.year, 3, 31)
+ last_october = datetime.datetime(date.year, 10, 31)
+ last_sunday_march = last_march - datetime.timedelta(days=last_march.isoweekday() % 7)
+ last_sunday_october = last_october - datetime.timedelta(days=last_october.isoweekday() % 7)
+ return last_sunday_march.replace(hour=2) <= date <= last_sunday_october.replace(hour=3)
+
+
+def rfc3339_to_atende(date):
+ date = datetime.datetime.fromisoformat(date)
+ date = date + datetime.timedelta(hours=1 if is_dst(date) else 0)
+ return int((date.timestamp() - 978307200) * 1000)
+
+
+class SejmIE(InfoExtractor):
+ _VALID_URL = (
+ r'https?://(?:www\.)?sejm\.gov\.pl/[Ss]ejm(?P<term>\d+)\.nsf/transmisje(?:_arch)?\.xsp(?:\?[^#]*)?#(?P<id>[\dA-F]+)',
+ r'https?://(?:www\.)?sejm\.gov\.pl/[Ss]ejm(?P<term>\d+)\.nsf/transmisje(?:_arch)?\.xsp\?(?:[^#]+&)?unid=(?P<id>[\dA-F]+)',
+ r'https?://sejm-embed\.redcdn\.pl/[Ss]ejm(?P<term>\d+)\.nsf/VideoFrame\.xsp/(?P<id>[\dA-F]+)',
+ )
+ IE_NAME = 'sejm'
+
+ _TESTS = [{
+ # multiple cameras, polish SL iterpreter
+ 'url': 'https://www.sejm.gov.pl/Sejm10.nsf/transmisje_arch.xsp#6181EF1AD9CEEBB5C1258A6D006452B5',
+ 'info_dict': {
+ 'id': '6181EF1AD9CEEBB5C1258A6D006452B5',
+ 'title': '1. posiedzenie Sejmu X kadencji',
+ 'duration': 20145,
+ 'live_status': 'was_live',
+ 'location': 'Sala Posiedzeń',
+ },
+ 'playlist': [{
+ 'info_dict': {
+ 'id': 'ENC01-722340000000-722360145000',
+ 'ext': 'mp4',
+ 'duration': 20145,
+ 'title': '1. posiedzenie Sejmu X kadencji - ENC01',
+ 'live_status': 'was_live',
+ },
+ }, {
+ 'info_dict': {
+ 'id': 'ENC30-722340000000-722360145000',
+ 'ext': 'mp4',
+ 'duration': 20145,
+ 'title': '1. posiedzenie Sejmu X kadencji - ENC30',
+ 'live_status': 'was_live',
+ },
+ }, {
+ 'info_dict': {
+ 'id': 'ENC31-722340000000-722360145000',
+ 'ext': 'mp4',
+ 'duration': 20145,
+ 'title': '1. posiedzenie Sejmu X kadencji - ENC31',
+ 'live_status': 'was_live',
+ },
+ }, {
+ 'info_dict': {
+ 'id': 'ENC32-722340000000-722360145000',
+ 'ext': 'mp4',
+ 'duration': 20145,
+ 'title': '1. posiedzenie Sejmu X kadencji - ENC32',
+ 'live_status': 'was_live',
+ },
+ }, {
+ # sign lang interpreter
+ 'info_dict': {
+ 'id': 'Migacz-ENC01-1-722340000000-722360145000',
+ 'ext': 'mp4',
+ 'duration': 20145,
+ 'title': '1. posiedzenie Sejmu X kadencji - Migacz-ENC01',
+ 'live_status': 'was_live',
+ },
+ }],
+ }, {
+ 'url': 'https://www.sejm.gov.pl/Sejm8.nsf/transmisje.xsp?unid=9377A9D65518E9A5C125808E002E9FF2',
+ 'info_dict': {
+ 'id': '9377A9D65518E9A5C125808E002E9FF2',
+ 'title': 'Debata "Lepsza Polska: obywatelska"',
+ 'description': 'KP .Nowoczesna',
+ 'duration': 8770,
+ 'live_status': 'was_live',
+ 'location': 'sala kolumnowa im. Kazimierza Pużaka (bud. C-D)',
+ },
+ 'playlist': [{
+ 'info_dict': {
+ 'id': 'ENC08-1-503831270000-503840040000',
+ 'ext': 'mp4',
+ 'duration': 8770,
+ 'title': 'Debata "Lepsza Polska: obywatelska" - ENC08',
+ 'live_status': 'was_live',
+ },
+ }],
+ }, {
+ # 7th term is very special, since it does not use redcdn livx
+ 'url': 'https://www.sejm.gov.pl/sejm7.nsf/transmisje_arch.xsp?rok=2015&month=11#A6E6D475ECCC6FE5C1257EF90034817F',
+ 'info_dict': {
+ 'id': 'A6E6D475ECCC6FE5C1257EF90034817F',
+ 'title': 'Konferencja prasowa - Stanowisko SLD ws. składu nowego rządu',
+ 'description': 'SLD - Biuro Prasowe Klubu',
+ 'duration': 514,
+ 'location': 'sala 101/bud. C',
+ 'live_status': 'was_live',
+ },
+ 'playlist': [{
+ 'info_dict': {
+ 'id': 'A6E6D475ECCC6FE5C1257EF90034817F',
+ 'ext': 'mp4',
+ 'title': 'Konferencja prasowa - Stanowisko SLD ws. składu nowego rządu',
+ 'duration': 514,
+ },
+ }],
+ }, {
+ 'url': 'https://sejm-embed.redcdn.pl/Sejm10.nsf/VideoFrame.xsp/FED58EABB97FBD53C1258A7400386492',
+ 'only_matching': True,
+ }]
+
+ def _real_extract(self, url):
+ term, video_id = self._match_valid_url(url).group('term', 'id')
+ frame = self._download_webpage(
+ f'https://sejm-embed.redcdn.pl/Sejm{term}.nsf/VideoFrame.xsp/{video_id}',
+ video_id)
+ # despite it says "transmisje_arch", it works for live streams too!
+ data = self._download_json(
+ f'https://www.sejm.gov.pl/Sejm{term}.nsf/transmisje_arch.xsp/json/{video_id}',
+ video_id)
+ params = data['params']
+
+ title = strip_or_none(data.get('title'))
+
+ if data.get('status') == 'VIDEO_ENDED':
+ live_status = 'was_live'
+ elif data.get('status') == 'VIDEO_PLAYING':
+ live_status = 'is_live'
+ else:
+ live_status = None
+ self.report_warning(f'unknown status: {data.get("status")}')
+
+ start_time = rfc3339_to_atende(params['start'])
+ # current streams have a stop time of *expected* end of session, but actual times
+ # can change during the transmission. setting a stop_time would artificially
+ # end the stream at that time, while the session actually keeps going.
+ if live_status == 'was_live':
+ stop_time = rfc3339_to_atende(params['stop'])
+ duration = (stop_time - start_time) // 1000
+ else:
+ stop_time, duration = None, None
+
+ entries = []
+
+ def add_entry(file, legacy_file=False):
+ if not file:
+ return
+ file = self._proto_relative_url(file)
+ if not legacy_file:
+ file = update_url_query(file, {'startTime': start_time})
+ if stop_time is not None:
+ file = update_url_query(file, {'stopTime': stop_time})
+ stream_id = self._search_regex(r'/o2/sejm/([^/]+)/[^./]+\.livx', file, 'stream id')
+ common_info = {
+ 'url': file,
+ 'duration': duration,
+ }
+ if legacy_file:
+ entries.append({
+ **common_info,
+ 'id': video_id,
+ 'title': title,
+ })
+ else:
+ entries.append({
+ **common_info,
+ '_type': 'url_transparent',
+ 'ie_key': RedCDNLivxIE.ie_key(),
+ 'id': stream_id,
+ 'title': join_nonempty(title, stream_id, delim=' - '),
+ })
+
+ cameras = self._search_json(
+ r'var\s+cameras\s*=', frame, 'camera list', video_id,
+ contains_pattern=r'\[(?s:.+)\]', transform_source=js_to_json,
+ fatal=False) or []
+ for camera_file in traverse_obj(cameras, (..., 'file', {dict})):
+ if camera_file.get('flv'):
+ add_entry(camera_file['flv'])
+ elif camera_file.get('mp4'):
+ # this is only a thing in 7th term. no streams before, and starting 8th it's redcdn livx
+ add_entry(camera_file['mp4'], legacy_file=True)
+ else:
+ self.report_warning('Unknown camera stream type found')
+
+ if params.get('mig'):
+ add_entry(self._search_regex(r"var sliUrl\s*=\s*'([^']+)'", frame, 'sign language interpreter url', fatal=False))
+
+ return {
+ '_type': 'playlist',
+ 'entries': entries,
+ 'id': video_id,
+ 'title': title,
+ 'description': clean_html(data.get('desc')) or None,
+ 'duration': duration,
+ 'live_status': live_status,
+ 'location': strip_or_none(data.get('location')),
+ }