[yt-dlp.git] / yt_dlp / extractor / twitcasting.py

import base64
import itertools
import re

from .common import InfoExtractor
from ..dependencies import websockets
from ..utils import (
    ExtractorError,
    UserNotLive,
    clean_html,
    float_or_none,
    get_element_by_class,
    get_element_by_id,
    int_or_none,
    parse_duration,
    qualities,
    str_to_int,
    traverse_obj,
    try_get,
    unified_timestamp,
    urlencode_postdata,
    urljoin,
)


class TwitCastingIE(InfoExtractor):
    _VALID_URL = r'https?://(?:[^/?#]+\.)?twitcasting\.tv/(?P<uploader_id>[^/?#]+)/(?:movie|twplayer)/(?P<id>\d+)'
    _M3U8_HEADERS = {
        'Origin': 'https://twitcasting.tv',
        'Referer': 'https://twitcasting.tv/',
    }
    _TESTS = [{
        'url': 'https://twitcasting.tv/ivetesangalo/movie/2357609',
        'md5': '745243cad58c4681dc752490f7540d7f',
        'info_dict': {
            'id': '2357609',
            'ext': 'mp4',
            'title': 'Live #2357609',
            'uploader_id': 'ivetesangalo',
            'description': 'Twitter Oficial da cantora brasileira Ivete Sangalo.',
            'thumbnail': r're:^https?://.*\.jpg$',
            'upload_date': '20110822',
            'timestamp': 1313978424,
            'duration': 32,
            'view_count': int,
        },
        'params': {
            'skip_download': True,
        },
    }, {
        'url': 'https://twitcasting.tv/mttbernardini/movie/3689740',
        'info_dict': {
            'id': '3689740',
            'ext': 'mp4',
            'title': 'Live playing something #3689740',
            'uploader_id': 'mttbernardini',
            'description': 'md5:1dc7efa2f1ab932fcd119265cebeec69',
            'thumbnail': r're:^https?://.*\.jpg$',
            'upload_date': '20120211',
            'timestamp': 1328995624,
            'duration': 681,
            'view_count': int,
        },
        'params': {
            'skip_download': True,
            'videopassword': 'abc',
        },
    }, {
        'url': 'https://twitcasting.tv/loft_heaven/movie/685979292',
        'info_dict': {
            'id': '685979292',
            'ext': 'mp4',
            'title': '【無料配信】南波一海のhear/here “ナタリー望月哲さんに聞く編集と「渋谷系狂騒曲」”',
            'uploader_id': 'loft_heaven',
            'description': 'md5:3a0c7b53019df987ce545c935538bacf',
            'upload_date': '20210604',
            'timestamp': 1622802114,
            'thumbnail': r're:^https?://.*\.jpg$',
            'duration': 6964,
            'view_count': int,
        },
        'params': {
            'skip_download': True,
        },
    }]

    def _parse_data_movie_playlist(self, dmp, video_id):
        # attempt 1: parse as JSON directly
        try:
            return self._parse_json(dmp, video_id)
        except ExtractorError:
            pass
        # attempt 2: decode reversed base64
        decoded = base64.b64decode(dmp[::-1])
        return self._parse_json(decoded, video_id)

    def _real_extract(self, url):
        uploader_id, video_id = self._match_valid_url(url).groups()

        webpage, urlh = self._download_webpage_handle(url, video_id)
        video_password = self.get_param('videopassword')
        request_data = None
        if video_password:
            request_data = urlencode_postdata({
                'password': video_password,
                **self._hidden_inputs(webpage),
            }, encoding='utf-8')
            webpage, urlh = self._download_webpage_handle(
                url, video_id, data=request_data,
                headers={'Origin': 'https://twitcasting.tv'},
                note='Trying video password')
        if urlh.url != url and request_data:
            webpage = self._download_webpage(
                urlh.url, video_id, data=request_data,
                headers={'Origin': 'https://twitcasting.tv'},
                note='Retrying authentication')
        # has to check here as the first request can contain password input form even if the password is correct
        if re.search(r'<form\s+method="POST">\s*<input\s+[^>]+?name="password"', webpage):
            raise ExtractorError('This video is protected by a password, use the --video-password option', expected=True)

        title = (clean_html(get_element_by_id('movietitle', webpage))
                 or self._html_search_meta(['og:title', 'twitter:title'], webpage, fatal=True))

        video_js_data = try_get(
            webpage,
            lambda x: self._parse_data_movie_playlist(self._search_regex(
                r'data-movie-playlist=\'([^\']+?)\'',
                x, 'movie playlist', default=None), video_id)['2'], list)

        thumbnail = traverse_obj(video_js_data, (0, 'thumbnailUrl')) or self._og_search_thumbnail(webpage)
        description = clean_html(get_element_by_id(
            'authorcomment', webpage)) or self._html_search_meta(
            ['description', 'og:description', 'twitter:description'], webpage)
        duration = (try_get(video_js_data, lambda x: sum(float_or_none(y.get('duration')) for y in x) / 1000)
                    or parse_duration(clean_html(get_element_by_class('tw-player-duration-time', webpage))))
        view_count = str_to_int(self._search_regex(
            (r'Total\s*:\s*Views\s*([\d,]+)', r'総視聴者\s*:\s*([\d,]+)\s*</'), webpage, 'views', None))
        timestamp = unified_timestamp(self._search_regex(
            r'data-toggle="true"[^>]+datetime="([^"]+)"',
            webpage, 'datetime', None))

        stream_server_data = self._download_json(
            'https://twitcasting.tv/streamserver.php?target=%s&mode=client' % uploader_id, video_id,
            'Downloading live info', fatal=False)

        is_live = any(f'data-{x}' in webpage for x in ['is-onlive="true"', 'live-type="live"', 'status="online"'])
        if not traverse_obj(stream_server_data, 'llfmp4') and is_live:
            self.raise_login_required(method='cookies')

        base_dict = {
            'title': title,
            'description': description,
            'thumbnail': thumbnail,
            'timestamp': timestamp,
            'uploader_id': uploader_id,
            'duration': duration,
            'view_count': view_count,
            'is_live': is_live,
        }

        def find_dmu(x):
            data_movie_url = self._search_regex(
                r'data-movie-url=(["\'])(?P<url>(?:(?!\1).)+)\1',
                x, 'm3u8 url', group='url', default=None)
            if data_movie_url:
                return [data_movie_url]

        m3u8_urls = (try_get(webpage, find_dmu, list)
                     or traverse_obj(video_js_data, (..., 'source', 'url'))
                     or ([f'https://twitcasting.tv/{uploader_id}/metastream.m3u8'] if is_live else None))
        if not m3u8_urls:
            raise ExtractorError('Failed to get m3u8 playlist')

        if is_live:
            m3u8_url = m3u8_urls[0]
            formats = self._extract_m3u8_formats(
                m3u8_url, video_id, ext='mp4', m3u8_id='hls',
                live=True, headers=self._M3U8_HEADERS)

            if traverse_obj(stream_server_data, ('hls', 'source')):
                formats.extend(self._extract_m3u8_formats(
                    m3u8_url, video_id, ext='mp4', m3u8_id='source',
                    live=True, query={'mode': 'source'},
                    note='Downloading source quality m3u8',
                    headers=self._M3U8_HEADERS, fatal=False))

            if websockets:
                qq = qualities(['base', 'mobilesource', 'main'])
                streams = traverse_obj(stream_server_data, ('llfmp4', 'streams')) or {}
                for mode, ws_url in streams.items():
                    formats.append({
                        'url': ws_url,
                        'format_id': 'ws-%s' % mode,
                        'ext': 'mp4',
                        'quality': qq(mode),
                        'source_preference': -10,
                        # TwitCasting simply sends moof atom directly over WS
                        'protocol': 'websocket_frag',
                    })

            infodict = {
                'formats': formats,
                '_format_sort_fields': ('source', ),
            }
        elif len(m3u8_urls) == 1:
            formats = self._extract_m3u8_formats(
                m3u8_urls[0], video_id, 'mp4', headers=self._M3U8_HEADERS)
            infodict = {
                # No problem here since there's only one manifest
                'formats': formats,
                'http_headers': self._M3U8_HEADERS,
            }
        else:
            infodict = {
                '_type': 'multi_video',
                'entries': [{
                    'id': f'{video_id}-{num}',
                    'url': m3u8_url,
                    'ext': 'mp4',
                    # Requesting the manifests here will cause download to fail.
                    # So use ffmpeg instead. See: https://github.com/yt-dlp/yt-dlp/issues/382
                    'protocol': 'm3u8',
                    'http_headers': self._M3U8_HEADERS,
                    **base_dict,
                } for (num, m3u8_url) in enumerate(m3u8_urls)],
            }

        return {
            'id': video_id,
            **base_dict,
            **infodict,
        }


class TwitCastingLiveIE(InfoExtractor):
    _VALID_URL = r'https?://(?:[^/?#]+\.)?twitcasting\.tv/(?P<id>[^/?#]+)/?(?:[#?]|$)'
    _TESTS = [{
        'url': 'https://twitcasting.tv/ivetesangalo',
        'only_matching': True,
    }, {
        'url': 'https://twitcasting.tv/c:unusedlive',
        'expected_exception': 'UserNotLive',
    }]

    _PROTECTED_LIVE_RE = r'(?s)(<span\s*class="tw-movie-thumbnail2-badge"\s*data-status="live">\s*LIVE)'

    def _real_extract(self, url):
        uploader_id = self._match_id(url)
        self.to_screen(
            'Downloading live video of user {0}. '
            'Pass "https://twitcasting.tv/{0}/show" to download the history'.format(uploader_id))

        webpage = self._download_webpage(url, uploader_id)
        is_live = self._search_regex(  # first pattern is for public live
            (r'(data-is-onlive="true")', self._PROTECTED_LIVE_RE), webpage, 'is live?', default=None)
        current_live = int_or_none(self._search_regex(
            (r'data-type="movie" data-id="(\d+)">',  # not available?
             r'tw-sound-flag-open-link" data-id="(\d+)" style=',  # not available?
             r'data-movie-id="(\d+)"'),  # if not currently live, value may be 0
            webpage, 'current live ID', default=None))
        if is_live and not current_live:
            # fetch unfiltered /show to find running livestreams; we can't get ID of the password-protected livestream above
            webpage = self._download_webpage(
                f'https://twitcasting.tv/{uploader_id}/show/', uploader_id,
                note='Downloading live history')
            is_live = self._search_regex(self._PROTECTED_LIVE_RE, webpage, 'is live?', default=None)
            if is_live:
                # get the first live; running live is always at the first
                current_live = self._search_regex(
                    r'(?s)<a\s+class="tw-movie-thumbnail2"\s*href="/[^/]+/movie/(?P<video_id>\d+)"\s*>.+?</a>',
                    webpage, 'current live ID 2', default=None, group='video_id')
        if not current_live:
            raise UserNotLive(video_id=uploader_id)
        return self.url_result(f'https://twitcasting.tv/{uploader_id}/movie/{current_live}', TwitCastingIE)


class TwitCastingUserIE(InfoExtractor):
    _VALID_URL = r'https?://(?:[^/?#]+\.)?twitcasting\.tv/(?P<id>[^/?#]+)/(:?show|archive)/?(?:[#?]|$)'
    _TESTS = [{
        'url': 'https://twitcasting.tv/natsuiromatsuri/archive/',
        'info_dict': {
            'id': 'natsuiromatsuri',
            'title': 'natsuiromatsuri - Live History',
        },
        'playlist_mincount': 235,
    }, {
        'url': 'https://twitcasting.tv/noriyukicas/show',
        'only_matching': True,
    }]

    def _entries(self, uploader_id):
        base_url = next_url = 'https://twitcasting.tv/%s/show' % uploader_id
        for page_num in itertools.count(1):
            webpage = self._download_webpage(
                next_url, uploader_id, query={'filter': 'watchable'}, note='Downloading page %d' % page_num)
            matches = re.finditer(
                r'''(?isx)<a\s+class="tw-movie-thumbnail"\s*href="(?P<url>/[^/]+/movie/\d+)"\s*>.+?</a>''',
                webpage)
            for mobj in matches:
                yield self.url_result(urljoin(base_url, mobj.group('url')))

            next_url = self._search_regex(
                r'<a href="(/%s/show/%d-\d+)[?"]' % (re.escape(uploader_id), page_num),
                webpage, 'next url', default=None)
            next_url = urljoin(base_url, next_url)
            if not next_url:
                return

    def _real_extract(self, url):
        uploader_id = self._match_id(url)
        return self.playlist_result(
            self._entries(uploader_id), uploader_id, '%s - Live History' % uploader_id)
Commit	Line	Data
da9a60ca	1	import base64
e85a3971	2	import itertools
036f9051	3	import re
036f9051	4
29f7c58a	5	from .common import InfoExtractor
9b8ee23b	6	from ..dependencies import websockets
29f7c58a	7	from ..utils import (
a4a42602	8	ExtractorError,
c1d71d0d AW	9	UserNotLive,
c1d71d0d AW	10	clean_html,
29f7c58a	11	float_or_none,
	12	get_element_by_class,
	13	get_element_by_id,
2325d03a	14	int_or_none,
29f7c58a	15	parse_duration,
e6779b94	16	qualities,
29f7c58a	17	str_to_int,
fabb27fc	18	traverse_obj,
e85a3971	19	try_get,
29f7c58a	20	unified_timestamp,
29f7c58a	21	urlencode_postdata,
e85a3971	22	urljoin,
29f7c58a	23	)
29f7c58a	24
036f9051	25
cf0db4d9	26	class TwitCastingIE(InfoExtractor):
cebbd33b	27	_VALID_URL = r'https?://(?:[^/?#]+\.)?twitcasting\.tv/(?P<uploader_id>[^/?#]+)/(?:movie\|twplayer)/(?P<id>\d+)'
a4a42602 LTHD	28	_M3U8_HEADERS = {
	29	'Origin': 'https://twitcasting.tv',
	30	'Referer': 'https://twitcasting.tv/',
	31	}
88b54749	32	_TESTS = [{
036f9051	33	'url': 'https://twitcasting.tv/ivetesangalo/movie/2357609',
	34	'md5': '745243cad58c4681dc752490f7540d7f',
	35	'info_dict': {
	36	'id': '2357609',
	37	'ext': 'mp4',
00a9a25c	38	'title': 'Live #2357609',
036f9051	39	'uploader_id': 'ivetesangalo',
29f7c58a	40	'description': 'Twitter Oficial da cantora brasileira Ivete Sangalo.',
036f9051	41	'thumbnail': r're:^https?://.*\.jpg$',
29f7c58a	42	'upload_date': '20110822',
9a9006ba	43	'timestamp': 1313978424,
29f7c58a	44	'duration': 32,
29f7c58a	45	'view_count': int,
cf0db4d9 S	46	},
	47	'params': {
	48	'skip_download': True,
	49	},
88b54749 MZ	50	}, {
	51	'url': 'https://twitcasting.tv/mttbernardini/movie/3689740',
	52	'info_dict': {
	53	'id': '3689740',
	54	'ext': 'mp4',
	55	'title': 'Live playing something #3689740',
	56	'uploader_id': 'mttbernardini',
9a9006ba	57	'description': 'md5:1dc7efa2f1ab932fcd119265cebeec69',
88b54749	58	'thumbnail': r're:^https?://.*\.jpg$',
9a9006ba S	59	'upload_date': '20120211',
9a9006ba S	60	'timestamp': 1328995624,
29f7c58a	61	'duration': 681,
29f7c58a	62	'view_count': int,
88b54749 MZ	63	},
	64	'params': {
	65	'skip_download': True,
	66	'videopassword': 'abc',
	67	},
a4a42602	68	}, {
a4a42602 LTHD	69	'url': 'https://twitcasting.tv/loft_heaven/movie/685979292',
	70	'info_dict': {
	71	'id': '685979292',
	72	'ext': 'mp4',
9a9006ba S	73	'title': '【無料配信】南波一海のhear/here “ナタリー望月哲さんに聞く編集と「渋谷系狂騒曲」”',
	74	'uploader_id': 'loft_heaven',
	75	'description': 'md5:3a0c7b53019df987ce545c935538bacf',
	76	'upload_date': '20210604',
	77	'timestamp': 1622802114,
	78	'thumbnail': r're:^https?://.*\.jpg$',
	79	'duration': 6964,
	80	'view_count': int,
	81	},
	82	'params': {
	83	'skip_download': True,
a4a42602	84	},
88b54749	85	}]
036f9051	86
da9a60ca L	87	def _parse_data_movie_playlist(self, dmp, video_id):
	88	# attempt 1: parse as JSON directly
	89	try:
	90	return self._parse_json(dmp, video_id)
	91	except ExtractorError:
	92	pass
	93	# attempt 2: decode reversed base64
	94	decoded = base64.b64decode(dmp[::-1])
	95	return self._parse_json(decoded, video_id)
	96
036f9051	97	def _real_extract(self, url):
5ad28e7f	98	uploader_id, video_id = self._match_valid_url(url).groups()
036f9051	99
9a9006ba	100	webpage, urlh = self._download_webpage_handle(url, video_id)
a06916d9	101	video_password = self.get_param('videopassword')
88b54749 MZ	102	request_data = None
	103	if video_password:
	104	request_data = urlencode_postdata({
	105	'password': video_password,
9a9006ba	106	**self._hidden_inputs(webpage),
5a13fdd2	107	}, encoding='utf-8')
9a9006ba S	108	webpage, urlh = self._download_webpage_handle(
	109	url, video_id, data=request_data,
	110	headers={'Origin': 'https://twitcasting.tv'},
	111	note='Trying video password')
3d2623a8	112	if urlh.url != url and request_data:
df635a09	113	webpage = self._download_webpage(
3d2623a8	114	urlh.url, video_id, data=request_data,
df635a09 LNO	115	headers={'Origin': 'https://twitcasting.tv'},
df635a09 LNO	116	note='Retrying authentication')
f099df14 LNO	117	# has to check here as the first request can contain password input form even if the password is correct
	118	if re.search(r'<form\s+method="POST">\s*<input\s+[^>]+?name="password"', webpage):
	119	raise ExtractorError('This video is protected by a password, use the --video-password option', expected=True)
036f9051	120
e85a3971	121	title = (clean_html(get_element_by_id('movietitle', webpage))
e85a3971	122	or self._html_search_meta(['og:title', 'twitter:title'], webpage, fatal=True))
cf0db4d9	123
a4a42602 LTHD	124	video_js_data = try_get(
a4a42602 LTHD	125	webpage,
da9a60ca	126	lambda x: self._parse_data_movie_playlist(self._search_regex(
a4a42602 LTHD	127	r'data-movie-playlist=\'([^\']+?)\'',
a4a42602 LTHD	128	x, 'movie playlist', default=None), video_id)['2'], list)
fabb27fc	129
a4a42602	130	thumbnail = traverse_obj(video_js_data, (0, 'thumbnailUrl')) or self._og_search_thumbnail(webpage)
29f7c58a	131	description = clean_html(get_element_by_id(
	132	'authorcomment', webpage)) or self._html_search_meta(
	133	['description', 'og:description', 'twitter:description'], webpage)
a4a42602 LTHD	134	duration = (try_get(video_js_data, lambda x: sum(float_or_none(y.get('duration')) for y in x) / 1000)
a4a42602 LTHD	135	or parse_duration(clean_html(get_element_by_class('tw-player-duration-time', webpage))))
29f7c58a	136	view_count = str_to_int(self._search_regex(
9a9006ba	137	(r'Total\s:\sViews\s([\d,]+)', r'総視聴者\s:\s([\d,]+)\s</'), webpage, 'views', None))
29f7c58a	138	timestamp = unified_timestamp(self._search_regex(
	139	r'data-toggle="true"[^>]+datetime="([^"]+)"',
	140	webpage, 'datetime', None))
cf0db4d9	141
a4a42602 LTHD	142	stream_server_data = self._download_json(
	143	'https://twitcasting.tv/streamserver.php?target=%s&mode=client' % uploader_id, video_id,
	144	'Downloading live info', fatal=False)
e85a3971	145
7b8b1cf5	146	is_live = any(f'data-{x}' in webpage for x in ['is-onlive="true"', 'live-type="live"', 'status="online"'])
a4a42602 LTHD	147	if not traverse_obj(stream_server_data, 'llfmp4') and is_live:
	148	self.raise_login_required(method='cookies')
	149
	150	base_dict = {
036f9051	151	'title': title,
	152	'description': description,
	153	'thumbnail': thumbnail,
29f7c58a	154	'timestamp': timestamp,
036f9051	155	'uploader_id': uploader_id,
29f7c58a	156	'duration': duration,
29f7c58a	157	'view_count': view_count,
e85a3971	158	'is_live': is_live,
036f9051	159	}
e85a3971	160
a4a42602 LTHD	161	def find_dmu(x):
	162	data_movie_url = self._search_regex(
	163	r'data-movie-url=(["\'])(?P<url>(?:(?!\1).)+)\1',
	164	x, 'm3u8 url', group='url', default=None)
	165	if data_movie_url:
	166	return [data_movie_url]
	167
	168	m3u8_urls = (try_get(webpage, find_dmu, list)
	169	or traverse_obj(video_js_data, (..., 'source', 'url'))
	170	or ([f'https://twitcasting.tv/{uploader_id}/metastream.m3u8'] if is_live else None))
	171	if not m3u8_urls:
	172	raise ExtractorError('Failed to get m3u8 playlist')
	173
	174	if is_live:
	175	m3u8_url = m3u8_urls[0]
	176	formats = self._extract_m3u8_formats(
	177	m3u8_url, video_id, ext='mp4', m3u8_id='hls',
	178	live=True, headers=self._M3U8_HEADERS)
	179
df635a09 LNO	180	if traverse_obj(stream_server_data, ('hls', 'source')):
	181	formats.extend(self._extract_m3u8_formats(
	182	m3u8_url, video_id, ext='mp4', m3u8_id='source',
	183	live=True, query={'mode': 'source'},
	184	note='Downloading source quality m3u8',
	185	headers=self._M3U8_HEADERS, fatal=False))
a4a42602	186
9b8ee23b	187	if websockets:
a4a42602 LTHD	188	qq = qualities(['base', 'mobilesource', 'main'])
	189	streams = traverse_obj(stream_server_data, ('llfmp4', 'streams')) or {}
	190	for mode, ws_url in streams.items():
	191	formats.append({
	192	'url': ws_url,
	193	'format_id': 'ws-%s' % mode,
	194	'ext': 'mp4',
	195	'quality': qq(mode),
df635a09	196	'source_preference': -10,
a4a42602 LTHD	197	# TwitCasting simply sends moof atom directly over WS
	198	'protocol': 'websocket_frag',
	199	})
	200
a4a42602	201	infodict = {
9f14daf2	202	'formats': formats,
9f14daf2	203	'_format_sort_fields': ('source', ),
a4a42602	204	}
06b1628d LNO	205	elif len(m3u8_urls) == 1:
	206	formats = self._extract_m3u8_formats(
	207	m3u8_urls[0], video_id, 'mp4', headers=self._M3U8_HEADERS)
06b1628d LNO	208	infodict = {
	209	# No problem here since there's only one manifest
	210	'formats': formats,
997378f9	211	'http_headers': self._M3U8_HEADERS,
06b1628d	212	}
a4a42602 LTHD	213	else:
	214	infodict = {
	215	'_type': 'multi_video',
	216	'entries': [{
	217	'id': f'{video_id}-{num}',
	218	'url': m3u8_url,
	219	'ext': 'mp4',
	220	# Requesting the manifests here will cause download to fail.
	221	# So use ffmpeg instead. See: https://github.com/yt-dlp/yt-dlp/issues/382
	222	'protocol': 'm3u8',
	223	'http_headers': self._M3U8_HEADERS,
	224	**base_dict,
	225	} for (num, m3u8_url) in enumerate(m3u8_urls)],
	226	}
	227
	228	return {
	229	'id': video_id,
	230	**base_dict,
	231	**infodict,
	232	}
	233
e85a3971	234
e85a3971	235	class TwitCastingLiveIE(InfoExtractor):
cebbd33b	236	_VALID_URL = r'https?://(?:[^/?#]+\.)?twitcasting\.tv/(?P<id>[^/?#]+)/?(?:[#?]\|$)'
e85a3971	237	_TESTS = [{
	238	'url': 'https://twitcasting.tv/ivetesangalo',
	239	'only_matching': True,
c1d71d0d AW	240	}, {
	241	'url': 'https://twitcasting.tv/c:unusedlive',
	242	'expected_exception': 'UserNotLive',
e85a3971	243	}]
e85a3971	244
2325d03a JC	245	_PROTECTED_LIVE_RE = r'(?s)(<span\sclass="tw-movie-thumbnail2-badge"\sdata-status="live">\s*LIVE)'
2325d03a JC	246
e85a3971	247	def _real_extract(self, url):
	248	uploader_id = self._match_id(url)
	249	self.to_screen(
	250	'Downloading live video of user {0}. '
	251	'Pass "https://twitcasting.tv/{0}/show" to download the history'.format(uploader_id))
	252
	253	webpage = self._download_webpage(url, uploader_id)
2325d03a JC	254	is_live = self._search_regex( # first pattern is for public live
	255	(r'(data-is-onlive="true")', self._PROTECTED_LIVE_RE), webpage, 'is live?', default=None)
	256	current_live = int_or_none(self._search_regex(
	257	(r'data-type="movie" data-id="(\d+)">', # not available?
	258	r'tw-sound-flag-open-link" data-id="(\d+)" style=', # not available?
	259	r'data-movie-id="(\d+)"'), # if not currently live, value may be 0
	260	webpage, 'current live ID', default=None))
	261	if is_live and not current_live:
07ea0014 LNO	262	# fetch unfiltered /show to find running livestreams; we can't get ID of the password-protected livestream above
	263	webpage = self._download_webpage(
	264	f'https://twitcasting.tv/{uploader_id}/show/', uploader_id,
	265	note='Downloading live history')
2325d03a	266	is_live = self._search_regex(self._PROTECTED_LIVE_RE, webpage, 'is live?', default=None)
07ea0014 LNO	267	if is_live:
	268	# get the first live; running live is always at the first
	269	current_live = self._search_regex(
2325d03a	270	r'(?s)<a\s+class="tw-movie-thumbnail2"\shref="/[^/]+/movie/(?P<video_id>\d+)"\s>.+?</a>',
07ea0014	271	webpage, 'current live ID 2', default=None, group='video_id')
e85a3971	272	if not current_live:
c1d71d0d	273	raise UserNotLive(video_id=uploader_id)
2325d03a	274	return self.url_result(f'https://twitcasting.tv/{uploader_id}/movie/{current_live}', TwitCastingIE)
e85a3971	275
	276
	277	class TwitCastingUserIE(InfoExtractor):
cebbd33b	278	_VALID_URL = r'https?://(?:[^/?#]+\.)?twitcasting\.tv/(?P<id>[^/?#]+)/(:?show\|archive)/?(?:[#?]\|$)'
e85a3971	279	_TESTS = [{
cebbd33b	280	'url': 'https://twitcasting.tv/natsuiromatsuri/archive/',
	281	'info_dict': {
	282	'id': 'natsuiromatsuri',
	283	'title': 'natsuiromatsuri - Live History',
	284	},
	285	'playlist_mincount': 235,
	286	}, {
e85a3971	287	'url': 'https://twitcasting.tv/noriyukicas/show',
	288	'only_matching': True,
	289	}]
	290
	291	def _entries(self, uploader_id):
	292	base_url = next_url = 'https://twitcasting.tv/%s/show' % uploader_id
	293	for page_num in itertools.count(1):
	294	webpage = self._download_webpage(
	295	next_url, uploader_id, query={'filter': 'watchable'}, note='Downloading page %d' % page_num)
	296	matches = re.finditer(
	297	r'''(?isx)<a\s+class="tw-movie-thumbnail"\shref="(?P<url>/[^/]+/movie/\d+)"\s>.+?</a>''',
	298	webpage)
	299	for mobj in matches:
	300	yield self.url_result(urljoin(base_url, mobj.group('url')))
	301
	302	next_url = self._search_regex(
	303	r'<a href="(/%s/show/%d-\d+)[?"]' % (re.escape(uploader_id), page_num),
	304	webpage, 'next url', default=None)
	305	next_url = urljoin(base_url, next_url)
	306	if not next_url:
	307	return
	308
	309	def _real_extract(self, url):
	310	uploader_id = self._match_id(url)
	311	return self.playlist_result(
	312	self._entries(uploader_id), uploader_id, '%s - Live History' % uploader_id)