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

from __future__ import unicode_literals

import re

from .common import InfoExtractor
from ..utils import (
    fix_xml_ampersands,
    parse_duration,
    qualities,
    strip_jsonp,
    unified_strdate,
)


class NPOBaseIE(InfoExtractor):
    def _get_token(self, video_id):
        token_page = self._download_webpage(
            'http://ida.omroep.nl/npoplayer/i.js',
            video_id, note='Downloading token')
        token = self._search_regex(
            r'npoplayer\.token = "(.+?)"', token_page, 'token')
        # Decryption algorithm extracted from http://npoplayer.omroep.nl/csjs/npoplayer-min.js
        token_l = list(token)
        first = second = None
        for i in range(5, len(token_l) - 4):
            if token_l[i].isdigit():
                if first is None:
                    first = i
                elif second is None:
                    second = i
        if first is None or second is None:
            first = 12
            second = 13

        token_l[first], token_l[second] = token_l[second], token_l[first]

        return ''.join(token_l)


class NPOIE(NPOBaseIE):
    IE_NAME = 'npo'
    IE_DESC = 'npo.nl and ntr.nl'
    _VALID_URL = r'''(?x)
                    (?:
                        npo:|
                        https?://
                            (?:www\.)?
                            (?:
                                npo\.nl/(?!live|radio)(?:[^/]+/){2}|
                                ntr\.nl/(?:[^/]+/){2,}|
                                omroepwnl\.nl/video/fragment/[^/]+__
                            )
                        )
                        (?P<id>[^/?#]+)
                '''

    _TESTS = [
        {
            'url': 'http://www.npo.nl/nieuwsuur/22-06-2014/VPWON_1220719',
            'md5': '4b3f9c429157ec4775f2c9cb7b911016',
            'info_dict': {
                'id': 'VPWON_1220719',
                'ext': 'm4v',
                'title': 'Nieuwsuur',
                'description': 'Dagelijks tussen tien en elf: nieuws, sport en achtergronden.',
                'upload_date': '20140622',
            },
        },
        {
            'url': 'http://www.npo.nl/de-mega-mike-mega-thomas-show/27-02-2009/VARA_101191800',
            'md5': 'da50a5787dbfc1603c4ad80f31c5120b',
            'info_dict': {
                'id': 'VARA_101191800',
                'ext': 'm4v',
                'title': 'De Mega Mike & Mega Thomas show',
                'description': 'md5:3b74c97fc9d6901d5a665aac0e5400f4',
                'upload_date': '20090227',
                'duration': 2400,
            },
        },
        {
            'url': 'http://www.npo.nl/tegenlicht/25-02-2013/VPWON_1169289',
            'md5': 'f8065e4e5a7824068ed3c7e783178f2c',
            'info_dict': {
                'id': 'VPWON_1169289',
                'ext': 'm4v',
                'title': 'Tegenlicht',
                'description': 'md5:52cf4eefbc96fffcbdc06d024147abea',
                'upload_date': '20130225',
                'duration': 3000,
            },
        },
        {
            'url': 'http://www.npo.nl/de-nieuwe-mens-deel-1/21-07-2010/WO_VPRO_043706',
            'info_dict': {
                'id': 'WO_VPRO_043706',
                'ext': 'wmv',
                'title': 'De nieuwe mens - Deel 1',
                'description': 'md5:518ae51ba1293ffb80d8d8ce90b74e4b',
                'duration': 4680,
            },
            'params': {
                # mplayer mms download
                'skip_download': True,
            }
        },
        # non asf in streams
        {
            'url': 'http://www.npo.nl/hoe-gaat-europa-verder-na-parijs/10-01-2015/WO_NOS_762771',
            'md5': 'b3da13de374cbe2d5332a7e910bef97f',
            'info_dict': {
                'id': 'WO_NOS_762771',
                'ext': 'mp4',
                'title': 'Hoe gaat Europa verder na Parijs?',
            },
        },
        {
            'url': 'http://www.ntr.nl/Aap-Poot-Pies/27/detail/Aap-poot-pies/VPWON_1233944#content',
            'md5': '01c6a2841675995da1f0cf776f03a9c3',
            'info_dict': {
                'id': 'VPWON_1233944',
                'ext': 'm4v',
                'title': 'Aap, poot, pies',
                'description': 'md5:c9c8005d1869ae65b858e82c01a91fde',
                'upload_date': '20150508',
                'duration': 599,
            },
        },
        {
            'url': 'http://www.omroepwnl.nl/video/fragment/vandaag-de-dag-verkiezingen__POMS_WNL_853698',
            'md5': 'd30cd8417b8b9bca1fdff27428860d08',
            'info_dict': {
                'id': 'POW_00996502',
                'ext': 'm4v',
                'title': '''"Dit is wel een 'landslide'..."''',
                'description': 'md5:f8d66d537dfb641380226e31ca57b8e8',
                'upload_date': '20150508',
                'duration': 462,
            },
        }
    ]

    def _real_extract(self, url):
        video_id = self._match_id(url)
        return self._get_info(video_id)

    def _get_info(self, video_id):
        metadata = self._download_json(
            'http://e.omroep.nl/metadata/%s' % video_id,
            video_id,
            # We have to remove the javascript callback
            transform_source=strip_jsonp,
        )

        # For some videos actual video id (prid) is different (e.g. for
        # http://www.omroepwnl.nl/video/fragment/vandaag-de-dag-verkiezingen__POMS_WNL_853698
        # video id is POMS_WNL_853698 but prid is POW_00996502)
        video_id = metadata.get('prid') or video_id

        token = self._get_token(video_id)

        formats = []

        pubopties = metadata.get('pubopties')
        if pubopties:
            quality = qualities(['adaptive', 'wmv_sb', 'h264_sb', 'wmv_bb', 'h264_bb', 'wvc1_std', 'h264_std'])
            for format_id in pubopties:
                format_info = self._download_json(
                    'http://ida.omroep.nl/odi/?prid=%s&puboptions=%s&adaptive=yes&token=%s'
                    % (video_id, format_id, token),
                    video_id, 'Downloading %s JSON' % format_id)
                if format_info.get('error_code', 0) or format_info.get('errorcode', 0):
                    continue
                streams = format_info.get('streams')
                if streams:
                    video_info = self._download_json(
                        streams[0] + '&type=json',
                        video_id, 'Downloading %s stream JSON' % format_id)
                else:
                    video_info = format_info
                video_url = video_info.get('url')
                if not video_url:
                    continue
                if format_id == 'adaptive':
                    formats.extend(self._extract_m3u8_formats(video_url, video_id))
                else:
                    formats.append({
                        'url': video_url,
                        'format_id': format_id,
                        'quality': quality(format_id),
                    })

        streams = metadata.get('streams')
        if streams:
            for i, stream in enumerate(streams):
                stream_url = stream.get('url')
                if not stream_url:
                    continue
                if '.asf' not in stream_url:
                    formats.append({
                        'url': stream_url,
                        'quality': stream.get('kwaliteit'),
                    })
                    continue
                asx = self._download_xml(
                    stream_url, video_id,
                    'Downloading stream %d ASX playlist' % i,
                    transform_source=fix_xml_ampersands)
                ref = asx.find('./ENTRY/Ref')
                if ref is None:
                    continue
                video_url = ref.get('href')
                if not video_url:
                    continue
                formats.append({
                    'url': video_url,
                    'ext': stream.get('formaat', 'asf'),
                    'quality': stream.get('kwaliteit'),
                })

        self._sort_formats(formats)

        subtitles = {}
        if metadata.get('tt888') == 'ja':
            subtitles['nl'] = [{
                'ext': 'vtt',
                'url': 'http://e.omroep.nl/tt888/%s' % video_id,
            }]

        return {
            'id': video_id,
            # prefer aflevering_titel if any since titel may be too generic, e.g.
            # http://tegenlicht.vpro.nl/afleveringen/2014-2015/access-to-africa.html
            'title': metadata.get('aflevering_titel') or metadata['titel'],
            'description': metadata.get('info'),
            'thumbnail': metadata.get('images', [{'url': None}])[-1]['url'],
            'upload_date': unified_strdate(metadata.get('gidsdatum')),
            'duration': parse_duration(metadata.get('tijdsduur')),
            'formats': formats,
            'subtitles': subtitles,
        }


class NPOLiveIE(NPOBaseIE):
    IE_NAME = 'npo.nl:live'
    _VALID_URL = r'https?://(?:www\.)?npo\.nl/live/(?P<id>.+)'

    _TEST = {
        'url': 'http://www.npo.nl/live/npo-1',
        'info_dict': {
            'id': 'LI_NEDERLAND1_136692',
            'display_id': 'npo-1',
            'ext': 'mp4',
            'title': 're:^Nederland 1 [0-9]{4}-[0-9]{2}-[0-9]{2} [0-9]{2}:[0-9]{2}$',
            'description': 'Livestream',
            'is_live': True,
        },
        'params': {
            'skip_download': True,
        }
    }

    def _real_extract(self, url):
        display_id = self._match_id(url)

        webpage = self._download_webpage(url, display_id)

        live_id = self._search_regex(
            r'data-prid="([^"]+)"', webpage, 'live id')

        metadata = self._download_json(
            'http://e.omroep.nl/metadata/%s' % live_id,
            display_id, transform_source=strip_jsonp)

        token = self._get_token(display_id)

        formats = []

        streams = metadata.get('streams')
        if streams:
            for stream in streams:
                stream_type = stream.get('type').lower()
                # smooth streaming is not supported
                if stream_type in ['ss', 'ms']:
                    continue
                stream_info = self._download_json(
                    'http://ida.omroep.nl/aapi/?stream=%s&token=%s&type=jsonp'
                    % (stream.get('url'), token),
                    display_id, 'Downloading %s JSON' % stream_type)
                if stream_info.get('error_code', 0) or stream_info.get('errorcode', 0):
                    continue
                stream_url = self._download_json(
                    stream_info['stream'], display_id,
                    'Downloading %s URL' % stream_type,
                    'Unable to download %s URL' % stream_type,
                    transform_source=strip_jsonp, fatal=False)
                if not stream_url:
                    continue
                if stream_type == 'hds':
                    f4m_formats = self._extract_f4m_formats(stream_url, display_id)
                    # f4m downloader downloads only piece of live stream
                    for f4m_format in f4m_formats:
                        f4m_format['preference'] = -1
                    formats.extend(f4m_formats)
                elif stream_type == 'hls':
                    formats.extend(self._extract_m3u8_formats(stream_url, display_id, 'mp4'))
                else:
                    formats.append({
                        'url': stream_url,
                        'preference': -10,
                    })

        self._sort_formats(formats)

        return {
            'id': live_id,
            'display_id': display_id,
            'title': self._live_title(metadata['titel']),
            'description': metadata['info'],
            'thumbnail': metadata.get('images', [{'url': None}])[-1]['url'],
            'formats': formats,
            'is_live': True,
        }


class NPORadioIE(InfoExtractor):
    IE_NAME = 'npo.nl:radio'
    _VALID_URL = r'https?://(?:www\.)?npo\.nl/radio/(?P<id>[^/]+)/?$'

    _TEST = {
        'url': 'http://www.npo.nl/radio/radio-1',
        'info_dict': {
            'id': 'radio-1',
            'ext': 'mp3',
            'title': 're:^NPO Radio 1 [0-9]{4}-[0-9]{2}-[0-9]{2} [0-9]{2}:[0-9]{2}$',
            'is_live': True,
        },
        'params': {
            'skip_download': True,
        }
    }

    @staticmethod
    def _html_get_attribute_regex(attribute):
        return r'{0}\s*=\s*\'([^\']+)\''.format(attribute)

    def _real_extract(self, url):
        video_id = self._match_id(url)

        webpage = self._download_webpage(url, video_id)

        title = self._html_search_regex(
            self._html_get_attribute_regex('data-channel'), webpage, 'title')

        stream = self._parse_json(
            self._html_search_regex(self._html_get_attribute_regex('data-streams'), webpage, 'data-streams'),
            video_id)

        codec = stream.get('codec')

        return {
            'id': video_id,
            'url': stream['url'],
            'title': self._live_title(title),
            'acodec': codec,
            'ext': codec,
            'is_live': True,
        }


class NPORadioFragmentIE(InfoExtractor):
    IE_NAME = 'npo.nl:radio:fragment'
    _VALID_URL = r'https?://(?:www\.)?npo\.nl/radio/[^/]+/fragment/(?P<id>\d+)'

    _TEST = {
        'url': 'http://www.npo.nl/radio/radio-5/fragment/174356',
        'md5': 'dd8cc470dad764d0fdc70a9a1e2d18c2',
        'info_dict': {
            'id': '174356',
            'ext': 'mp3',
            'title': 'Jubileumconcert Willeke Alberti',
        },
    }

    def _real_extract(self, url):
        audio_id = self._match_id(url)

        webpage = self._download_webpage(url, audio_id)

        title = self._html_search_regex(
            r'href="/radio/[^/]+/fragment/%s" title="([^"]+)"' % audio_id,
            webpage, 'title')

        audio_url = self._search_regex(
            r"data-streams='([^']+)'", webpage, 'audio url')

        return {
            'id': audio_id,
            'url': audio_url,
            'title': title,
        }


class VPROIE(NPOIE):
    _VALID_URL = r'https?://(?:www\.)?(?:tegenlicht\.)?vpro\.nl/(?:[^/]+/){2,}(?P<id>[^/]+)\.html'

    _TESTS = [
        {
            'url': 'http://tegenlicht.vpro.nl/afleveringen/2012-2013/de-toekomst-komt-uit-afrika.html',
            'md5': 'f8065e4e5a7824068ed3c7e783178f2c',
            'info_dict': {
                'id': 'VPWON_1169289',
                'ext': 'm4v',
                'title': 'De toekomst komt uit Afrika',
                'description': 'md5:52cf4eefbc96fffcbdc06d024147abea',
                'upload_date': '20130225',
            },
        },
        {
            'url': 'http://www.vpro.nl/programmas/2doc/2015/sergio-herman.html',
            'info_dict': {
                'id': 'sergio-herman',
                'title': 'Sergio Herman: Fucking perfect',
            },
            'playlist_count': 2,
        },
        {
            # playlist with youtube embed
            'url': 'http://www.vpro.nl/programmas/2doc/2015/education-education.html',
            'info_dict': {
                'id': 'education-education',
                'title': '2Doc',
            },
            'playlist_count': 2,
        }
    ]

    def _real_extract(self, url):
        playlist_id = self._match_id(url)

        webpage = self._download_webpage(url, playlist_id)

        entries = [
            self.url_result('npo:%s' % video_id if not video_id.startswith('http') else video_id)
            for video_id in re.findall(r'data-media-id="([^"]+)"', webpage)
        ]

        playlist_title = self._search_regex(
            r'<title>\s*([^>]+?)\s*-\s*Teledoc\s*-\s*VPRO\s*</title>',
            webpage, 'playlist title', default=None) or self._og_search_title(webpage)

        return self.playlist_result(entries, playlist_id, playlist_title)


class WNLIE(InfoExtractor):
    _VALID_URL = r'https?://(?:www\.)?omroepwnl\.nl/video/detail/(?P<id>[^/]+)__\d+'

    _TEST = {
        'url': 'http://www.omroepwnl.nl/video/detail/vandaag-de-dag-6-mei__060515',
        'info_dict': {
            'id': 'vandaag-de-dag-6-mei',
            'title': 'Vandaag de Dag 6 mei',
        },
        'playlist_count': 4,
    }

    def _real_extract(self, url):
        playlist_id = self._match_id(url)

        webpage = self._download_webpage(url, playlist_id)

        entries = [
            self.url_result('npo:%s' % video_id, 'NPO')
            for video_id, part in re.findall(
                r'<a[^>]+href="([^"]+)"[^>]+class="js-mid"[^>]*>(Deel \d+)', webpage)
        ]

        playlist_title = self._html_search_regex(
            r'(?s)<h1[^>]+class="subject"[^>]*>(.+?)</h1>',
            webpage, 'playlist title')

        return self.playlist_result(entries, playlist_id, playlist_title)
Commit	Line	Data
331ae266 JMF	1	from __future__ import unicode_literals
331ae266 JMF	2
525daedd S	3	import re
525daedd S	4
171ca612	5	from .common import InfoExtractor
331ae266	6	from ..utils import (
5c4a81d9	7	fix_xml_ampersands,
944a3de2	8	parse_duration,
c9ea760e	9	qualities,
609a61e3	10	strip_jsonp,
5c4a81d9	11	unified_strdate,
331ae266 JMF	12	)
	13
	14
b9b42f2e	15	class NPOBaseIE(InfoExtractor):
04e0bac2 S	16	def _get_token(self, video_id):
	17	token_page = self._download_webpage(
	18	'http://ida.omroep.nl/npoplayer/i.js',
	19	video_id, note='Downloading token')
8d06a624	20	token = self._search_regex(
04e0bac2	21	r'npoplayer\.token = "(.+?)"', token_page, 'token')
f5ca97e3	22	# Decryption algorithm extracted from http://npoplayer.omroep.nl/csjs/npoplayer-min.js
8d06a624 YCH	23	token_l = list(token)
	24	first = second = None
	25	for i in range(5, len(token_l) - 4):
	26	if token_l[i].isdigit():
	27	if first is None:
	28	first = i
	29	elif second is None:
	30	second = i
	31	if first is None or second is None:
	32	first = 12
	33	second = 13
	34
	35	token_l[first], token_l[second] = token_l[second], token_l[first]
	36
	37	return ''.join(token_l)
04e0bac2 S	38
	39
	40	class NPOIE(NPOBaseIE):
e118031e S	41	IE_NAME = 'npo'
e118031e S	42	IE_DESC = 'npo.nl and ntr.nl'
525daedd	43	_VALID_URL = r'''(?x)
03f32a7e S	44	(?:
	45	npo:\|
	46	https?://
	47	(?:www\.)?
	48	(?:
	49	npo\.nl/(?!live\|radio)(?:[^/]+/){2}\|
	50	ntr\.nl/(?:[^/]+/){2,}\|
	51	omroepwnl\.nl/video/fragment/[^/]+__
	52	)
525daedd S	53	)
	54	(?P<id>[^/?#]+)
	55	'''
331ae266	56
944a3de2 S	57	_TESTS = [
	58	{
	59	'url': 'http://www.npo.nl/nieuwsuur/22-06-2014/VPWON_1220719',
	60	'md5': '4b3f9c429157ec4775f2c9cb7b911016',
	61	'info_dict': {
	62	'id': 'VPWON_1220719',
	63	'ext': 'm4v',
	64	'title': 'Nieuwsuur',
	65	'description': 'Dagelijks tussen tien en elf: nieuws, sport en achtergronden.',
	66	'upload_date': '20140622',
	67	},
331ae266	68	},
944a3de2 S	69	{
	70	'url': 'http://www.npo.nl/de-mega-mike-mega-thomas-show/27-02-2009/VARA_101191800',
	71	'md5': 'da50a5787dbfc1603c4ad80f31c5120b',
	72	'info_dict': {
	73	'id': 'VARA_101191800',
	74	'ext': 'm4v',
	75	'title': 'De Mega Mike & Mega Thomas show',
	76	'description': 'md5:3b74c97fc9d6901d5a665aac0e5400f4',
	77	'upload_date': '20090227',
	78	'duration': 2400,
	79	},
	80	},
	81	{
	82	'url': 'http://www.npo.nl/tegenlicht/25-02-2013/VPWON_1169289',
	83	'md5': 'f8065e4e5a7824068ed3c7e783178f2c',
	84	'info_dict': {
	85	'id': 'VPWON_1169289',
	86	'ext': 'm4v',
	87	'title': 'Tegenlicht',
50ea2bb2	88	'description': 'md5:52cf4eefbc96fffcbdc06d024147abea',
944a3de2 S	89	'upload_date': '20130225',
	90	'duration': 3000,
	91	},
1ff30d7b S	92	},
	93	{
	94	'url': 'http://www.npo.nl/de-nieuwe-mens-deel-1/21-07-2010/WO_VPRO_043706',
	95	'info_dict': {
	96	'id': 'WO_VPRO_043706',
	97	'ext': 'wmv',
	98	'title': 'De nieuwe mens - Deel 1',
	99	'description': 'md5:518ae51ba1293ffb80d8d8ce90b74e4b',
	100	'duration': 4680,
	101	},
	102	'params': {
	103	# mplayer mms download
	104	'skip_download': True,
	105	}
	106	},
c85f3683 S	107	# non asf in streams
	108	{
	109	'url': 'http://www.npo.nl/hoe-gaat-europa-verder-na-parijs/10-01-2015/WO_NOS_762771',
	110	'md5': 'b3da13de374cbe2d5332a7e910bef97f',
	111	'info_dict': {
	112	'id': 'WO_NOS_762771',
	113	'ext': 'mp4',
	114	'title': 'Hoe gaat Europa verder na Parijs?',
	115	},
	116	},
e118031e S	117	{
	118	'url': 'http://www.ntr.nl/Aap-Poot-Pies/27/detail/Aap-poot-pies/VPWON_1233944#content',
	119	'md5': '01c6a2841675995da1f0cf776f03a9c3',
	120	'info_dict': {
	121	'id': 'VPWON_1233944',
	122	'ext': 'm4v',
	123	'title': 'Aap, poot, pies',
	124	'description': 'md5:c9c8005d1869ae65b858e82c01a91fde',
	125	'upload_date': '20150508',
	126	'duration': 599,
	127	},
525daedd S	128	},
	129	{
	130	'url': 'http://www.omroepwnl.nl/video/fragment/vandaag-de-dag-verkiezingen__POMS_WNL_853698',
	131	'md5': 'd30cd8417b8b9bca1fdff27428860d08',
	132	'info_dict': {
	133	'id': 'POW_00996502',
	134	'ext': 'm4v',
	135	'title': '''"Dit is wel een 'landslide'..."''',
	136	'description': 'md5:f8d66d537dfb641380226e31ca57b8e8',
	137	'upload_date': '20150508',
	138	'duration': 462,
	139	},
e118031e	140	}
944a3de2	141	]
331ae266 JMF	142
331ae266 JMF	143	def _real_extract(self, url):
04e0bac2	144	video_id = self._match_id(url)
d0df9292	145	return self._get_info(video_id)
331ae266	146
d0df9292	147	def _get_info(self, video_id):
331ae266	148	metadata = self._download_json(
8d06a624	149	'http://e.omroep.nl/metadata/%s' % video_id,
331ae266 JMF	150	video_id,
331ae266 JMF	151	# We have to remove the javascript callback
609a61e3	152	transform_source=strip_jsonp,
331ae266	153	)
04e0bac2	154
525daedd S	155	# For some videos actual video id (prid) is different (e.g. for
	156	# http://www.omroepwnl.nl/video/fragment/vandaag-de-dag-verkiezingen__POMS_WNL_853698
	157	# video id is POMS_WNL_853698 but prid is POW_00996502)
	158	video_id = metadata.get('prid') or video_id
	159
04e0bac2	160	token = self._get_token(video_id)
331ae266	161
c9ea760e	162	formats = []
1ff30d7b S	163
	164	pubopties = metadata.get('pubopties')
	165	if pubopties:
	166	quality = qualities(['adaptive', 'wmv_sb', 'h264_sb', 'wmv_bb', 'h264_bb', 'wvc1_std', 'h264_std'])
	167	for format_id in pubopties:
	168	format_info = self._download_json(
	169	'http://ida.omroep.nl/odi/?prid=%s&puboptions=%s&adaptive=yes&token=%s'
	170	% (video_id, format_id, token),
	171	video_id, 'Downloading %s JSON' % format_id)
	172	if format_info.get('error_code', 0) or format_info.get('errorcode', 0):
	173	continue
	174	streams = format_info.get('streams')
	175	if streams:
	176	video_info = self._download_json(
	177	streams[0] + '&type=json',
	178	video_id, 'Downloading %s stream JSON' % format_id)
	179	else:
	180	video_info = format_info
	181	video_url = video_info.get('url')
	182	if not video_url:
	183	continue
	184	if format_id == 'adaptive':
	185	formats.extend(self._extract_m3u8_formats(video_url, video_id))
	186	else:
	187	formats.append({
	188	'url': video_url,
	189	'format_id': format_id,
	190	'quality': quality(format_id),
	191	})
	192
	193	streams = metadata.get('streams')
	194	if streams:
	195	for i, stream in enumerate(streams):
	196	stream_url = stream.get('url')
	197	if not stream_url:
	198	continue
5c4a81d9	199	if '.asf' not in stream_url:
a0977064 S	200	formats.append({
	201	'url': stream_url,
	202	'quality': stream.get('kwaliteit'),
	203	})
	204	continue
1ff30d7b S	205	asx = self._download_xml(
	206	stream_url, video_id,
	207	'Downloading stream %d ASX playlist' % i,
	208	transform_source=fix_xml_ampersands)
	209	ref = asx.find('./ENTRY/Ref')
	210	if ref is None:
	211	continue
	212	video_url = ref.get('href')
	213	if not video_url:
	214	continue
c9ea760e	215	formats.append({
944a3de2	216	'url': video_url,
1ff30d7b S	217	'ext': stream.get('formaat', 'asf'),
1ff30d7b S	218	'quality': stream.get('kwaliteit'),
c9ea760e	219	})
1ff30d7b	220
9e91449c	221	self._sort_formats(formats)
25e5ebf3	222
9e91449c S	223	subtitles = {}
9e91449c S	224	if metadata.get('tt888') == 'ja':
b9b42f2e JMF	225	subtitles['nl'] = [{
	226	'ext': 'vtt',
	227	'url': 'http://e.omroep.nl/tt888/%s' % video_id,
	228	}]
331ae266 JMF	229
	230	return {
	231	'id': video_id,
5ba761eb S	232	# prefer aflevering_titel if any since titel may be too generic, e.g.
	233	# http://tegenlicht.vpro.nl/afleveringen/2014-2015/access-to-africa.html
	234	'title': metadata.get('aflevering_titel') or metadata['titel'],
536b0700	235	'description': metadata.get('info'),
944a3de2 S	236	'thumbnail': metadata.get('images', [{'url': None}])[-1]['url'],
	237	'upload_date': unified_strdate(metadata.get('gidsdatum')),
	238	'duration': parse_duration(metadata.get('tijdsduur')),
c9ea760e	239	'formats': formats,
25e5ebf3	240	'subtitles': subtitles,
331ae266	241	}
d0df9292 JMF	242
d0df9292 JMF	243
04e0bac2 S	244	class NPOLiveIE(NPOBaseIE):
04e0bac2 S	245	IE_NAME = 'npo.nl:live'
171ca612	246	_VALID_URL = r'https?://(?:www\.)?npo\.nl/live/(?P<id>.+)'
04e0bac2 S	247
	248	_TEST = {
	249	'url': 'http://www.npo.nl/live/npo-1',
	250	'info_dict': {
	251	'id': 'LI_NEDERLAND1_136692',
	252	'display_id': 'npo-1',
	253	'ext': 'mp4',
	254	'title': 're:^Nederland 1 [0-9]{4}-[0-9]{2}-[0-9]{2} [0-9]{2}:[0-9]{2}$',
	255	'description': 'Livestream',
	256	'is_live': True,
	257	},
	258	'params': {
	259	'skip_download': True,
	260	}
	261	}
	262
	263	def _real_extract(self, url):
	264	display_id = self._match_id(url)
	265
	266	webpage = self._download_webpage(url, display_id)
	267
	268	live_id = self._search_regex(
	269	r'data-prid="([^"]+)"', webpage, 'live id')
	270
	271	metadata = self._download_json(
	272	'http://e.omroep.nl/metadata/%s' % live_id,
	273	display_id, transform_source=strip_jsonp)
	274
	275	token = self._get_token(display_id)
	276
	277	formats = []
	278
	279	streams = metadata.get('streams')
	280	if streams:
	281	for stream in streams:
	282	stream_type = stream.get('type').lower()
32aaeca7 S	283	# smooth streaming is not supported
32aaeca7 S	284	if stream_type in ['ss', 'ms']:
04e0bac2 S	285	continue
	286	stream_info = self._download_json(
	287	'http://ida.omroep.nl/aapi/?stream=%s&token=%s&type=jsonp'
	288	% (stream.get('url'), token),
	289	display_id, 'Downloading %s JSON' % stream_type)
	290	if stream_info.get('error_code', 0) or stream_info.get('errorcode', 0):
	291	continue
	292	stream_url = self._download_json(
	293	stream_info['stream'], display_id,
	294	'Downloading %s URL' % stream_type,
b0872c19 S	295	'Unable to download %s URL' % stream_type,
	296	transform_source=strip_jsonp, fatal=False)
	297	if not stream_url:
	298	continue
04e0bac2 S	299	if stream_type == 'hds':
	300	f4m_formats = self._extract_f4m_formats(stream_url, display_id)
	301	# f4m downloader downloads only piece of live stream
	302	for f4m_format in f4m_formats:
	303	f4m_format['preference'] = -1
	304	formats.extend(f4m_formats)
	305	elif stream_type == 'hls':
	306	formats.extend(self._extract_m3u8_formats(stream_url, display_id, 'mp4'))
	307	else:
	308	formats.append({
	309	'url': stream_url,
32aaeca7	310	'preference': -10,
04e0bac2 S	311	})
	312
	313	self._sort_formats(formats)
	314
	315	return {
	316	'id': live_id,
	317	'display_id': display_id,
	318	'title': self._live_title(metadata['titel']),
	319	'description': metadata['info'],
	320	'thumbnail': metadata.get('images', [{'url': None}])[-1]['url'],
	321	'formats': formats,
	322	'is_live': True,
	323	}
	324
	325
171ca612 S	326	class NPORadioIE(InfoExtractor):
	327	IE_NAME = 'npo.nl:radio'
	328	_VALID_URL = r'https?://(?:www\.)?npo\.nl/radio/(?P<id>[^/]+)/?$'
	329
	330	_TEST = {
	331	'url': 'http://www.npo.nl/radio/radio-1',
	332	'info_dict': {
	333	'id': 'radio-1',
	334	'ext': 'mp3',
	335	'title': 're:^NPO Radio 1 [0-9]{4}-[0-9]{2}-[0-9]{2} [0-9]{2}:[0-9]{2}$',
	336	'is_live': True,
	337	},
	338	'params': {
	339	'skip_download': True,
	340	}
	341	}
	342
	343	@staticmethod
	344	def _html_get_attribute_regex(attribute):
	345	return r'{0}\s=\s\'([^\']+)\''.format(attribute)
	346
	347	def _real_extract(self, url):
	348	video_id = self._match_id(url)
	349
	350	webpage = self._download_webpage(url, video_id)
	351
	352	title = self._html_search_regex(
	353	self._html_get_attribute_regex('data-channel'), webpage, 'title')
	354
	355	stream = self._parse_json(
	356	self._html_search_regex(self._html_get_attribute_regex('data-streams'), webpage, 'data-streams'),
	357	video_id)
	358
	359	codec = stream.get('codec')
	360
	361	return {
	362	'id': video_id,
	363	'url': stream['url'],
	364	'title': self._live_title(title),
	365	'acodec': codec,
	366	'ext': codec,
	367	'is_live': True,
	368	}
	369
	370
	371	class NPORadioFragmentIE(InfoExtractor):
	372	IE_NAME = 'npo.nl:radio:fragment'
	373	_VALID_URL = r'https?://(?:www\.)?npo\.nl/radio/[^/]+/fragment/(?P<id>\d+)'
	374
	375	_TEST = {
	376	'url': 'http://www.npo.nl/radio/radio-5/fragment/174356',
	377	'md5': 'dd8cc470dad764d0fdc70a9a1e2d18c2',
	378	'info_dict': {
	379	'id': '174356',
	380	'ext': 'mp3',
	381	'title': 'Jubileumconcert Willeke Alberti',
	382	},
	383	}
	384
	385	def _real_extract(self, url):
	386	audio_id = self._match_id(url)
	387
	388	webpage = self._download_webpage(url, audio_id)
	389
390	title = self._html_search_regex(
391	r'href="/radio/[^/]+/fragment/%s" title="([^"]+)"' % audio_id,
392	webpage, 'title')
393
394	audio_url = self._search_regex(
395	r"data-streams='([^']+)'", webpage, 'audio url')
396
397	return {
398	'id': audio_id,
399	'url': audio_url,
400	'title': title,
401	}
402
403
611ac379 S	404	class VPROIE(NPOIE):
611ac379 S	405	_VALID_URL = r'https?://(?:www\.)?(?:tegenlicht\.)?vpro\.nl/(?:[^/]+/){2,}(?P<id>[^/]+)\.html'
d0df9292 JMF	406
	407	_TESTS = [
	408	{
	409	'url': 'http://tegenlicht.vpro.nl/afleveringen/2012-2013/de-toekomst-komt-uit-afrika.html',
	410	'md5': 'f8065e4e5a7824068ed3c7e783178f2c',
	411	'info_dict': {
	412	'id': 'VPWON_1169289',
	413	'ext': 'm4v',
5ba761eb	414	'title': 'De toekomst komt uit Afrika',
611ac379	415	'description': 'md5:52cf4eefbc96fffcbdc06d024147abea',
d0df9292 JMF	416	'upload_date': '20130225',
	417	},
	418	},
611ac379 S	419	{
	420	'url': 'http://www.vpro.nl/programmas/2doc/2015/sergio-herman.html',
	421	'info_dict': {
	422	'id': 'sergio-herman',
	423	'title': 'Sergio Herman: Fucking perfect',
	424	},
	425	'playlist_count': 2,
574f42d7 S	426	},
	427	{
	428	# playlist with youtube embed
	429	'url': 'http://www.vpro.nl/programmas/2doc/2015/education-education.html',
	430	'info_dict': {
	431	'id': 'education-education',
	432	'title': '2Doc',
	433	},
	434	'playlist_count': 2,
611ac379	435	}
d0df9292 JMF	436	]
	437
	438	def _real_extract(self, url):
611ac379 S	439	playlist_id = self._match_id(url)
	440
	441	webpage = self._download_webpage(url, playlist_id)
	442
	443	entries = [
574f42d7	444	self.url_result('npo:%s' % video_id if not video_id.startswith('http') else video_id)
611ac379 S	445	for video_id in re.findall(r'data-media-id="([^"]+)"', webpage)
	446	]
	447
	448	playlist_title = self._search_regex(
	449	r'<title>\s([^>]+?)\s-\sTeledoc\s-\sVPRO\s</title>',
	450	webpage, 'playlist title', default=None) or self._og_search_title(webpage)
	451
	452	return self.playlist_result(entries, playlist_id, playlist_title)
03f32a7e S	453
	454
	455	class WNLIE(InfoExtractor):
	456	_VALID_URL = r'https?://(?:www\.)?omroepwnl\.nl/video/detail/(?P<id>[^/]+)__\d+'
	457
	458	_TEST = {
	459	'url': 'http://www.omroepwnl.nl/video/detail/vandaag-de-dag-6-mei__060515',
	460	'info_dict': {
	461	'id': 'vandaag-de-dag-6-mei',
	462	'title': 'Vandaag de Dag 6 mei',
	463	},
	464	'playlist_count': 4,
	465	}
	466
	467	def _real_extract(self, url):
	468	playlist_id = self._match_id(url)
	469
	470	webpage = self._download_webpage(url, playlist_id)
	471
	472	entries = [
	473	self.url_result('npo:%s' % video_id, 'NPO')
	474	for video_id, part in re.findall(
	475	r'<a[^>]+href="([^"]+)"[^>]+class="js-mid"[^>]*>(Deel \d+)', webpage)
	476	]
	477
	478	playlist_title = self._html_search_regex(
	479	r'(?s)<h1[^>]+class="subject"[^>]*>(.+?)</h1>',
	480	webpage, 'playlist title')
	481
	482	return self.playlist_result(entries, playlist_id, playlist_title)