[yt-dlp.git] / youtube_dl / extractor / livestream.py

from __future__ import unicode_literals

import re
import json

from .common import InfoExtractor
from ..utils import (
    compat_str,
    compat_urllib_parse_urlparse,
    compat_urlparse,
    ExtractorError,
    find_xpath_attr,
    int_or_none,
    orderedSet,
    xpath_with_ns,
)


class LivestreamIE(InfoExtractor):
    IE_NAME = 'livestream'
    _VALID_URL = r'https?://new\.livestream\.com/.*?/(?P<event_name>.*?)(/videos/(?P<id>[0-9]+)(?:/player)?)?/?(?:$|[?#])'
    _TESTS = [{
        'url': 'http://new.livestream.com/CoheedandCambria/WebsterHall/videos/4719370',
        'md5': '53274c76ba7754fb0e8d072716f2292b',
        'info_dict': {
            'id': '4719370',
            'ext': 'mp4',
            'title': 'Live from Webster Hall NYC',
            'upload_date': '20121012',
            'like_count': int,
            'view_count': int,
            'thumbnail': 're:^http://.*\.jpg$'
        }
    }, {
        'url': 'http://new.livestream.com/tedx/cityenglish',
        'info_dict': {
            'title': 'TEDCity2.0 (English)',
        },
        'playlist_mincount': 4,
    }, {
        'url': 'https://new.livestream.com/accounts/362/events/3557232/videos/67864563/player?autoPlay=false&height=360&mute=false&width=640',
        'only_matching': True,
    }]

    def _parse_smil(self, video_id, smil_url):
        formats = []
        _SWITCH_XPATH = (
            './/{http://www.w3.org/2001/SMIL20/Language}body/'
            '{http://www.w3.org/2001/SMIL20/Language}switch')
        smil_doc = self._download_xml(
            smil_url, video_id,
            note='Downloading SMIL information',
            errnote='Unable to download SMIL information',
            fatal=False)
        if smil_doc is False:  # Download failed
            return formats
        title_node = find_xpath_attr(
            smil_doc, './/{http://www.w3.org/2001/SMIL20/Language}meta',
            'name', 'title')
        if title_node is None:
            self.report_warning('Cannot find SMIL id')
            switch_node = smil_doc.find(_SWITCH_XPATH)
        else:
            title_id = title_node.attrib['content']
            switch_node = find_xpath_attr(
                smil_doc, _SWITCH_XPATH, 'id', title_id)
        if switch_node is None:
            raise ExtractorError('Cannot find switch node')
        video_nodes = switch_node.findall(
            '{http://www.w3.org/2001/SMIL20/Language}video')

        for vn in video_nodes:
            tbr = int_or_none(vn.attrib.get('system-bitrate'))
            furl = (
                'http://livestream-f.akamaihd.net/%s?v=3.0.3&fp=WIN%%2014,0,0,145' %
                (vn.attrib['src']))
            if 'clipBegin' in vn.attrib:
                furl += '&ssek=' + vn.attrib['clipBegin']
            formats.append({
                'url': furl,
                'format_id': 'smil_%d' % tbr,
                'ext': 'flv',
                'tbr': tbr,
                'preference': -1000,
            })
        return formats

    def _extract_video_info(self, video_data):
        video_id = compat_str(video_data['id'])

        FORMAT_KEYS = (
            ('sd', 'progressive_url'),
            ('hd', 'progressive_url_hd'),
        )
        formats = [{
            'format_id': format_id,
            'url': video_data[key],
            'quality': i + 1,
        } for i, (format_id, key) in enumerate(FORMAT_KEYS)
            if video_data.get(key)]

        smil_url = video_data.get('smil_url')
        if smil_url:
            formats.extend(self._parse_smil(video_id, smil_url))
        self._sort_formats(formats)

        return {
            'id': video_id,
            'formats': formats,
            'title': video_data['caption'],
            'thumbnail': video_data.get('thumbnail_url'),
            'upload_date': video_data['updated_at'].replace('-', '')[:8],
            'like_count': video_data.get('likes', {}).get('total'),
            'view_count': video_data.get('views'),
        }

    def _real_extract(self, url):
        mobj = re.match(self._VALID_URL, url)
        video_id = mobj.group('id')
        event_name = mobj.group('event_name')
        webpage = self._download_webpage(url, video_id or event_name)

        og_video = self._og_search_video_url(
            webpage, 'player url', fatal=False, default=None)
        if og_video is not None:
            query_str = compat_urllib_parse_urlparse(og_video).query
            query = compat_urlparse.parse_qs(query_str)
            if 'play_url' in query:
                api_url = query['play_url'][0].replace('.smil', '')
                info = json.loads(self._download_webpage(
                    api_url, video_id, 'Downloading video info'))
                return self._extract_video_info(info)

        config_json = self._search_regex(
            r'window.config = ({.*?});', webpage, 'window config')
        info = json.loads(config_json)['event']

        def is_relevant(vdata, vid):
            result = vdata['type'] == 'video'
            if video_id is not None:
                result = result and compat_str(vdata['data']['id']) == vid
            return result

        videos = [self._extract_video_info(video_data['data'])
                  for video_data in info['feed']['data']
                  if is_relevant(video_data, video_id)]
        if video_id is None:
            # This is an event page:
            return self.playlist_result(videos, info['id'], info['full_name'])
        else:
            if not videos:
                raise ExtractorError('Cannot find video %s' % video_id)
            return videos[0]


# The original version of Livestream uses a different system
class LivestreamOriginalIE(InfoExtractor):
    IE_NAME = 'livestream:original'
    _VALID_URL = r'''(?x)https?://www\.livestream\.com/
        (?P<user>[^/]+)/(?P<type>video|folder)
        (?:\?.*?Id=|/)(?P<id>.*?)(&|$)
        '''
    _TESTS = [{
        'url': 'http://www.livestream.com/dealbook/video?clipId=pla_8aa4a3f1-ba15-46a4-893b-902210e138fb',
        'info_dict': {
            'id': 'pla_8aa4a3f1-ba15-46a4-893b-902210e138fb',
            'ext': 'flv',
            'title': 'Spark 1 (BitCoin) with Cameron Winklevoss & Tyler Winklevoss of Winklevoss Capital',
        },
        'params': {
            # rtmp
            'skip_download': True,
        },
    }, {
        'url': 'https://www.livestream.com/newplay/folder?dirId=a07bf706-d0e4-4e75-a747-b021d84f2fd3',
        'info_dict': {
            'id': 'a07bf706-d0e4-4e75-a747-b021d84f2fd3',
        },
        'playlist_mincount': 4,
    }]

    def _extract_video(self, user, video_id):
        api_url = 'http://x{0}x.api.channel.livestream.com/2.0/clipdetails?extendedInfo=true&id={1}'.format(user, video_id)

        info = self._download_xml(api_url, video_id)
        item = info.find('channel').find('item')
        ns = {'media': 'http://search.yahoo.com/mrss'}
        thumbnail_url = item.find(xpath_with_ns('media:thumbnail', ns)).attrib['url']
        # Remove the extension and number from the path (like 1.jpg)
        path = self._search_regex(r'(user-files/.+)_.*?\.jpg$', thumbnail_url, 'path')

        return {
            'id': video_id,
            'title': item.find('title').text,
            'url': 'rtmp://extondemand.livestream.com/ondemand',
            'play_path': 'trans/dv15/mogulus-{0}'.format(path),
            'player_url': 'http://static.livestream.com/chromelessPlayer/v21/playerapi.swf?hash=5uetk&v=0803&classid=D27CDB6E-AE6D-11cf-96B8-444553540000&jsEnabled=false&wmode=opaque',
            'ext': 'flv',
            'thumbnail': thumbnail_url,
        }

    def _extract_folder(self, url, folder_id):
        webpage = self._download_webpage(url, folder_id)
        paths = orderedSet(re.findall(
            r'''(?x)(?:
                <li\s+class="folder">\s*<a\s+href="|
                <a\s+href="(?=https?://livestre\.am/)
            )([^"]+)"''', webpage))

        return {
            '_type': 'playlist',
            'id': folder_id,
            'entries': [{
                '_type': 'url',
                'url': compat_urlparse.urljoin(url, p),
            } for p in paths],
        }

    def _real_extract(self, url):
        mobj = re.match(self._VALID_URL, url)
        id = mobj.group('id')
        user = mobj.group('user')
        url_type = mobj.group('type')
        if url_type == 'folder':
            return self._extract_folder(url, id)
        else:
            return self._extract_video(user, id)


# The server doesn't support HEAD request, the generic extractor can't detect
# the redirection
class LivestreamShortenerIE(InfoExtractor):
    IE_NAME = 'livestream:shortener'
    IE_DESC = False  # Do not list
    _VALID_URL = r'https?://livestre\.am/(?P<id>.+)'

    def _real_extract(self, url):
        mobj = re.match(self._VALID_URL, url)
        id = mobj.group('id')
        webpage = self._download_webpage(url, id)

        return {
            '_type': 'url',
            'url': self._og_search_url(webpage),
        }
Commit	Line	Data
c5469e04 S	1	from __future__ import unicode_literals
c5469e04 S	2
b4444d5c JMF	3	import re
	4	import json
	5
	6	from .common import InfoExtractor
b00ca882	7	from ..utils import (
cbf915f3	8	compat_str,
b00ca882 JMF	9	compat_urllib_parse_urlparse,
b00ca882 JMF	10	compat_urlparse,
cbf915f3 PH	11	ExtractorError,
	12	find_xpath_attr,
	13	int_or_none,
78338f71	14	orderedSet,
cbf915f3	15	xpath_with_ns,
b00ca882	16	)
b4444d5c JMF	17
	18
	19	class LivestreamIE(InfoExtractor):
c5469e04	20	IE_NAME = 'livestream'
af63fed7	21	_VALID_URL = r'https?://new\.livestream\.com/.?/(?P<event_name>.?)(/videos/(?P<id>[0-9]+)(?:/player)?)?/?(?:$\|[?#])'
22a6f150	22	_TESTS = [{
c5469e04 S	23	'url': 'http://new.livestream.com/CoheedandCambria/WebsterHall/videos/4719370',
	24	'md5': '53274c76ba7754fb0e8d072716f2292b',
	25	'info_dict': {
	26	'id': '4719370',
	27	'ext': 'mp4',
	28	'title': 'Live from Webster Hall NYC',
	29	'upload_date': '20121012',
cbf915f3 PH	30	'like_count': int,
	31	'view_count': int,
	32	'thumbnail': 're:^http://.*\.jpg$'
b4444d5c	33	}
22a6f150 PH	34	}, {
	35	'url': 'http://new.livestream.com/tedx/cityenglish',
	36	'info_dict': {
	37	'title': 'TEDCity2.0 (English)',
	38	},
	39	'playlist_mincount': 4,
af63fed7 PH	40	}, {
	41	'url': 'https://new.livestream.com/accounts/362/events/3557232/videos/67864563/player?autoPlay=false&height=360&mute=false&width=640',
	42	'only_matching': True,
22a6f150	43	}]
b4444d5c	44
8f3034d8 PH	45	def _parse_smil(self, video_id, smil_url):
	46	formats = []
	47	_SWITCH_XPATH = (
	48	'.//{http://www.w3.org/2001/SMIL20/Language}body/'
	49	'{http://www.w3.org/2001/SMIL20/Language}switch')
	50	smil_doc = self._download_xml(
	51	smil_url, video_id,
	52	note='Downloading SMIL information',
	53	errnote='Unable to download SMIL information',
	54	fatal=False)
	55	if smil_doc is False: # Download failed
	56	return formats
	57	title_node = find_xpath_attr(
	58	smil_doc, './/{http://www.w3.org/2001/SMIL20/Language}meta',
	59	'name', 'title')
	60	if title_node is None:
	61	self.report_warning('Cannot find SMIL id')
	62	switch_node = smil_doc.find(_SWITCH_XPATH)
	63	else:
	64	title_id = title_node.attrib['content']
	65	switch_node = find_xpath_attr(
	66	smil_doc, _SWITCH_XPATH, 'id', title_id)
	67	if switch_node is None:
	68	raise ExtractorError('Cannot find switch node')
	69	video_nodes = switch_node.findall(
	70	'{http://www.w3.org/2001/SMIL20/Language}video')
	71
	72	for vn in video_nodes:
	73	tbr = int_or_none(vn.attrib.get('system-bitrate'))
	74	furl = (
	75	'http://livestream-f.akamaihd.net/%s?v=3.0.3&fp=WIN%%2014,0,0,145' %
	76	(vn.attrib['src']))
	77	if 'clipBegin' in vn.attrib:
	78	furl += '&ssek=' + vn.attrib['clipBegin']
	79	formats.append({
	80	'url': furl,
	81	'format_id': 'smil_%d' % tbr,
	82	'ext': 'flv',
	83	'tbr': tbr,
	84	'preference': -1000,
	85	})
	86	return formats
	87
b4444d5c	88	def _extract_video_info(self, video_data):
cbf915f3 PH	89	video_id = compat_str(video_data['id'])
	90
	91	FORMAT_KEYS = (
	92	('sd', 'progressive_url'),
	93	('hd', 'progressive_url_hd'),
72e785f3	94	)
cbf915f3 PH	95	formats = [{
	96	'format_id': format_id,
	97	'url': video_data[key],
	98	'quality': i + 1,
	99	} for i, (format_id, key) in enumerate(FORMAT_KEYS)
	100	if video_data.get(key)]
	101
	102	smil_url = video_data.get('smil_url')
	103	if smil_url:
8f3034d8	104	formats.extend(self._parse_smil(video_id, smil_url))
cbf915f3 PH	105	self._sort_formats(formats)
cbf915f3 PH	106
c5469e04	107	return {
cbf915f3 PH	108	'id': video_id,
cbf915f3 PH	109	'formats': formats,
c5469e04	110	'title': video_data['caption'],
cbf915f3	111	'thumbnail': video_data.get('thumbnail_url'),
c5469e04	112	'upload_date': video_data['updated_at'].replace('-', '')[:8],
cbf915f3 PH	113	'like_count': video_data.get('likes', {}).get('total'),
cbf915f3 PH	114	'view_count': video_data.get('views'),
c5469e04	115	}
b4444d5c JMF	116
	117	def _real_extract(self, url):
	118	mobj = re.match(self._VALID_URL, url)
	119	video_id = mobj.group('id')
	120	event_name = mobj.group('event_name')
	121	webpage = self._download_webpage(url, video_id or event_name)
	122
22a6f150 PH	123	og_video = self._og_search_video_url(
	124	webpage, 'player url', fatal=False, default=None)
	125	if og_video is not None:
b4444d5c JMF	126	query_str = compat_urllib_parse_urlparse(og_video).query
b4444d5c JMF	127	query = compat_urlparse.parse_qs(query_str)
22a6f150 PH	128	if 'play_url' in query:
	129	api_url = query['play_url'][0].replace('.smil', '')
	130	info = json.loads(self._download_webpage(
	131	api_url, video_id, 'Downloading video info'))
	132	return self._extract_video_info(info)
	133
	134	config_json = self._search_regex(
	135	r'window.config = ({.*?});', webpage, 'window config')
	136	info = json.loads(config_json)['event']
	137
	138	def is_relevant(vdata, vid):
	139	result = vdata['type'] == 'video'
	140	if video_id is not None:
	141	result = result and compat_str(vdata['data']['id']) == vid
	142	return result
	143
	144	videos = [self._extract_video_info(video_data['data'])
	145	for video_data in info['feed']['data']
	146	if is_relevant(video_data, video_id)]
	147	if video_id is None:
	148	# This is an event page:
	149	return self.playlist_result(videos, info['id'], info['full_name'])
	150	else:
	151	if not videos:
	152	raise ExtractorError('Cannot find video %s' % video_id)
	153	return videos[0]
c66d2baa JMF	154
	155
	156	# The original version of Livestream uses a different system
	157	class LivestreamOriginalIE(InfoExtractor):
c5469e04	158	IE_NAME = 'livestream:original'
78338f71 JMF	159	_VALID_URL = r'''(?x)https?://www\.livestream\.com/
	160	(?P<user>[^/]+)/(?P<type>video\|folder)
	161	(?:\?.?Id=\|/)(?P<id>.?)(&\|$)
	162	'''
22a6f150	163	_TESTS = [{
c5469e04 S	164	'url': 'http://www.livestream.com/dealbook/video?clipId=pla_8aa4a3f1-ba15-46a4-893b-902210e138fb',
	165	'info_dict': {
	166	'id': 'pla_8aa4a3f1-ba15-46a4-893b-902210e138fb',
	167	'ext': 'flv',
	168	'title': 'Spark 1 (BitCoin) with Cameron Winklevoss & Tyler Winklevoss of Winklevoss Capital',
c66d2baa	169	},
c5469e04	170	'params': {
c66d2baa	171	# rtmp
c5469e04	172	'skip_download': True,
c66d2baa	173	},
22a6f150 PH	174	}, {
	175	'url': 'https://www.livestream.com/newplay/folder?dirId=a07bf706-d0e4-4e75-a747-b021d84f2fd3',
	176	'info_dict': {
	177	'id': 'a07bf706-d0e4-4e75-a747-b021d84f2fd3',
	178	},
	179	'playlist_mincount': 4,
	180	}]
c66d2baa	181
78338f71	182	def _extract_video(self, user, video_id):
c66d2baa JMF	183	api_url = 'http://x{0}x.api.channel.livestream.com/2.0/clipdetails?extendedInfo=true&id={1}'.format(user, video_id)
c66d2baa JMF	184
e26f8712	185	info = self._download_xml(api_url, video_id)
c66d2baa JMF	186	item = info.find('channel').find('item')
	187	ns = {'media': 'http://search.yahoo.com/mrss'}
	188	thumbnail_url = item.find(xpath_with_ns('media:thumbnail', ns)).attrib['url']
	189	# Remove the extension and number from the path (like 1.jpg)
c5469e04	190	path = self._search_regex(r'(user-files/.+)_.*?\.jpg$', thumbnail_url, 'path')
c66d2baa JMF	191
	192	return {
	193	'id': video_id,
	194	'title': item.find('title').text,
	195	'url': 'rtmp://extondemand.livestream.com/ondemand',
83855f3a PH	196	'play_path': 'trans/dv15/mogulus-{0}'.format(path),
83855f3a PH	197	'player_url': 'http://static.livestream.com/chromelessPlayer/v21/playerapi.swf?hash=5uetk&v=0803&classid=D27CDB6E-AE6D-11cf-96B8-444553540000&jsEnabled=false&wmode=opaque',
c66d2baa JMF	198	'ext': 'flv',
	199	'thumbnail': thumbnail_url,
	200	}
78338f71 JMF	201
	202	def _extract_folder(self, url, folder_id):
	203	webpage = self._download_webpage(url, folder_id)
22a6f150 PH	204	paths = orderedSet(re.findall(
	205	r'''(?x)(?:
	206	<li\s+class="folder">\s*<a\s+href="\|
	207	<a\s+href="(?=https?://livestre\.am/)
	208	)([^"]+)"''', webpage))
78338f71 JMF	209
	210	return {
	211	'_type': 'playlist',
	212	'id': folder_id,
	213	'entries': [{
	214	'_type': 'url',
22a6f150 PH	215	'url': compat_urlparse.urljoin(url, p),
22a6f150 PH	216	} for p in paths],
78338f71 JMF	217	}
	218
	219	def _real_extract(self, url):
	220	mobj = re.match(self._VALID_URL, url)
	221	id = mobj.group('id')
	222	user = mobj.group('user')
	223	url_type = mobj.group('type')
	224	if url_type == 'folder':
	225	return self._extract_folder(url, id)
	226	else:
	227	return self._extract_video(user, id)
	228
	229
	230	# The server doesn't support HEAD request, the generic extractor can't detect
	231	# the redirection
	232	class LivestreamShortenerIE(InfoExtractor):
	233	IE_NAME = 'livestream:shortener'
	234	IE_DESC = False # Do not list
	235	_VALID_URL = r'https?://livestre\.am/(?P<id>.+)'
	236
	237	def _real_extract(self, url):
	238	mobj = re.match(self._VALID_URL, url)
	239	id = mobj.group('id')
	240	webpage = self._download_webpage(url, id)
	241
	242	return {
	243	'_type': 'url',
	244	'url': self._og_search_url(webpage),
	245	}