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

# encoding: utf-8
from __future__ import unicode_literals

import re

from .common import InfoExtractor
from ..compat import (
    compat_parse_qs,
    compat_urllib_parse_urlparse,
)
from ..utils import (
    find_xpath_attr,
    unified_strdate,
    get_element_by_attribute,
    int_or_none,
    qualities,
)

# There are different sources of video in arte.tv, the extraction process
# is different for each one. The videos usually expire in 7 days, so we can't
# add tests.


class ArteTvIE(InfoExtractor):
    _VALID_URL = r'http://videos\.arte\.tv/(?P<lang>fr|de)/.*-(?P<id>.*?)\.html'
    IE_NAME = 'arte.tv'

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

        ref_xml_url = url.replace('/videos/', '/do_delegate/videos/')
        ref_xml_url = ref_xml_url.replace('.html', ',view,asPlayerXml.xml')
        ref_xml_doc = self._download_xml(
            ref_xml_url, video_id, note='Downloading metadata')
        config_node = find_xpath_attr(ref_xml_doc, './/video', 'lang', lang)
        config_xml_url = config_node.attrib['ref']
        config = self._download_xml(
            config_xml_url, video_id, note='Downloading configuration')

        formats = [{
            'format_id': q.attrib['quality'],
            # The playpath starts at 'mp4:', if we don't manually
            # split the url, rtmpdump will incorrectly parse them
            'url': q.text.split('mp4:', 1)[0],
            'play_path': 'mp4:' + q.text.split('mp4:', 1)[1],
            'ext': 'flv',
            'quality': 2 if q.attrib['quality'] == 'hd' else 1,
        } for q in config.findall('./urls/url')]
        self._sort_formats(formats)

        title = config.find('.//name').text
        thumbnail = config.find('.//firstThumbnailUrl').text
        return {
            'id': video_id,
            'title': title,
            'thumbnail': thumbnail,
            'formats': formats,
        }


class ArteTVPlus7IE(InfoExtractor):
    IE_NAME = 'arte.tv:+7'
    _VALID_URL = r'https?://(?:www\.)?arte\.tv/guide/(?P<lang>fr|de)/(?:(?:sendungen|emissions)/)?(?P<id>.*?)/(?P<name>.*?)(\?.*)?'

    @classmethod
    def _extract_url_info(cls, url):
        mobj = re.match(cls._VALID_URL, url)
        lang = mobj.group('lang')
        # This is not a real id, it can be for example AJT for the news
        # http://www.arte.tv/guide/fr/emissions/AJT/arte-journal
        video_id = mobj.group('id')
        return video_id, lang

    def _real_extract(self, url):
        video_id, lang = self._extract_url_info(url)
        webpage = self._download_webpage(url, video_id)
        return self._extract_from_webpage(webpage, video_id, lang)

    def _extract_from_webpage(self, webpage, video_id, lang):
        json_url = self._html_search_regex(
            [r'arte_vp_url=["\'](.*?)["\']', r'data-url=["\']([^"]+)["\']'],
            webpage, 'json vp url', default=None)
        if not json_url:
            iframe_url = self._html_search_regex(
                r'<iframe[^>]+src=(["\'])(?P<url>.+\bjson_url=.+?)\1',
                webpage, 'iframe url', group='url')
            json_url = compat_parse_qs(
                compat_urllib_parse_urlparse(iframe_url).query)['json_url'][0]
        return self._extract_from_json_url(json_url, video_id, lang)

    def _extract_from_json_url(self, json_url, video_id, lang):
        info = self._download_json(json_url, video_id)
        player_info = info['videoJsonPlayer']

        upload_date_str = player_info.get('shootingDate')
        if not upload_date_str:
            upload_date_str = player_info.get('VDA', '').split(' ')[0]

        title = player_info['VTI'].strip()
        subtitle = player_info.get('VSU', '').strip()
        if subtitle:
            title += ' - %s' % subtitle

        info_dict = {
            'id': player_info['VID'],
            'title': title,
            'description': player_info.get('VDE'),
            'upload_date': unified_strdate(upload_date_str),
            'thumbnail': player_info.get('programImage') or player_info.get('VTU', {}).get('IUR'),
        }
        qfunc = qualities(['HQ', 'MQ', 'EQ', 'SQ'])

        formats = []
        for format_id, format_dict in player_info['VSR'].items():
            f = dict(format_dict)
            versionCode = f.get('versionCode')

            langcode = {
                'fr': 'F',
                'de': 'A',
            }.get(lang, lang)
            lang_rexs = [r'VO?%s' % langcode, r'VO?.-ST%s' % langcode]
            lang_pref = (
                None if versionCode is None else (
                    10 if any(re.match(r, versionCode) for r in lang_rexs)
                    else -10))
            source_pref = 0
            if versionCode is not None:
                # The original version with subtitles has lower relevance
                if re.match(r'VO-ST(F|A)', versionCode):
                    source_pref -= 10
                # The version with sourds/mal subtitles has also lower relevance
                elif re.match(r'VO?(F|A)-STM\1', versionCode):
                    source_pref -= 9
            format = {
                'format_id': format_id,
                'preference': -10 if f.get('videoFormat') == 'M3U8' else None,
                'language_preference': lang_pref,
                'format_note': '%s, %s' % (f.get('versionCode'), f.get('versionLibelle')),
                'width': int_or_none(f.get('width')),
                'height': int_or_none(f.get('height')),
                'tbr': int_or_none(f.get('bitrate')),
                'quality': qfunc(f.get('quality')),
                'source_preference': source_pref,
            }

            if f.get('mediaType') == 'rtmp':
                format['url'] = f['streamer']
                format['play_path'] = 'mp4:' + f['url']
                format['ext'] = 'flv'
            else:
                format['url'] = f['url']

            formats.append(format)

        self._check_formats(formats, video_id)
        self._sort_formats(formats)

        info_dict['formats'] = formats
        return info_dict


# It also uses the arte_vp_url url from the webpage to extract the information
class ArteTVCreativeIE(ArteTVPlus7IE):
    IE_NAME = 'arte.tv:creative'
    _VALID_URL = r'https?://creative\.arte\.tv/(?P<lang>fr|de)/(?:magazine?/)?(?P<id>[^?#]+)'

    _TESTS = [{
        'url': 'http://creative.arte.tv/de/magazin/agentur-amateur-corporate-design',
        'info_dict': {
            'id': '72176',
            'ext': 'mp4',
            'title': 'Folge 2 - Corporate Design',
            'upload_date': '20131004',
        },
    }, {
        'url': 'http://creative.arte.tv/fr/Monty-Python-Reunion',
        'info_dict': {
            'id': '160676',
            'ext': 'mp4',
            'title': 'Monty Python live (mostly)',
            'description': 'Événement ! Quarante-cinq ans après leurs premiers succès, les légendaires Monty Python remontent sur scène.\n',
            'upload_date': '20140805',
        }
    }]


class ArteTVFutureIE(ArteTVPlus7IE):
    IE_NAME = 'arte.tv:future'
    _VALID_URL = r'https?://future\.arte\.tv/(?P<lang>fr|de)/(thema|sujet)/.*?#article-anchor-(?P<id>\d+)'

    _TEST = {
        'url': 'http://future.arte.tv/fr/sujet/info-sciences#article-anchor-7081',
        'info_dict': {
            'id': '5201',
            'ext': 'mp4',
            'title': 'Les champignons au secours de la planète',
            'upload_date': '20131101',
        },
    }

    def _real_extract(self, url):
        anchor_id, lang = self._extract_url_info(url)
        webpage = self._download_webpage(url, anchor_id)
        row = self._search_regex(
            r'(?s)id="%s"[^>]*>.+?(<div[^>]*arte_vp_url[^>]*>)' % anchor_id,
            webpage, 'row')
        return self._extract_from_webpage(row, anchor_id, lang)


class ArteTVDDCIE(ArteTVPlus7IE):
    IE_NAME = 'arte.tv:ddc'
    _VALID_URL = r'https?://ddc\.arte\.tv/(?P<lang>emission|folge)/(?P<id>.+)'

    def _real_extract(self, url):
        video_id, lang = self._extract_url_info(url)
        if lang == 'folge':
            lang = 'de'
        elif lang == 'emission':
            lang = 'fr'
        webpage = self._download_webpage(url, video_id)
        scriptElement = get_element_by_attribute('class', 'visu_video_block', webpage)
        script_url = self._html_search_regex(r'src="(.*?)"', scriptElement, 'script url')
        javascriptPlayerGenerator = self._download_webpage(script_url, video_id, 'Download javascript player generator')
        json_url = self._search_regex(r"json_url=(.*)&rendering_place.*", javascriptPlayerGenerator, 'json url')
        return self._extract_from_json_url(json_url, video_id, lang)


class ArteTVConcertIE(ArteTVPlus7IE):
    IE_NAME = 'arte.tv:concert'
    _VALID_URL = r'https?://concert\.arte\.tv/(?P<lang>de|fr)/(?P<id>.+)'

    _TEST = {
        'url': 'http://concert.arte.tv/de/notwist-im-pariser-konzertclub-divan-du-monde',
        'md5': '9ea035b7bd69696b67aa2ccaaa218161',
        'info_dict': {
            'id': '186',
            'ext': 'mp4',
            'title': 'The Notwist im Pariser Konzertclub "Divan du Monde"',
            'upload_date': '20140128',
            'description': 'md5:486eb08f991552ade77439fe6d82c305',
        },
    }


class ArteTVEmbedIE(ArteTVPlus7IE):
    IE_NAME = 'arte.tv:embed'
    _VALID_URL = r'''(?x)
        http://www\.arte\.tv
        /playerv2/embed\.php\?json_url=
        (?P<json_url>
            http://arte\.tv/papi/tvguide/videos/stream/player/
            (?P<lang>[^/]+)/(?P<id>[^/]+)[^&]*
        )
    '''

    def _real_extract(self, url):
        mobj = re.match(self._VALID_URL, url)
        video_id = mobj.group('id')
        lang = mobj.group('lang')
        json_url = mobj.group('json_url')
        return self._extract_from_json_url(json_url, video_id, lang)
Commit	Line	Data
69a0c470	1	# encoding: utf-8
3798eadc PH	2	from __future__ import unicode_literals
3798eadc PH	3
d5822b96	4	import re
d5822b96 PH	5
d5822b96 PH	6	from .common import InfoExtractor
393ca8c9 S	7	from ..compat import (
	8	compat_parse_qs,
	9	compat_urllib_parse_urlparse,
	10	)
d5822b96	11	from ..utils import (
df50a412	12	find_xpath_attr,
d5822b96	13	unified_strdate,
56a8ab7d	14	get_element_by_attribute,
d24a2b20	15	int_or_none,
aff2f4f4	16	qualities,
d5822b96 PH	17	)
d5822b96 PH	18
5f6a1245	19	# There are different sources of video in arte.tv, the extraction process
c40f5cf4 JMF	20	# is different for each one. The videos usually expire in 7 days, so we can't
	21	# add tests.
	22
d5822b96	23
515bbe4b	24	class ArteTvIE(InfoExtractor):
878d11ec	25	_VALID_URL = r'http://videos\.arte\.tv/(?P<lang>fr\|de)/.-(?P<id>.?)\.html'
3798eadc	26	IE_NAME = 'arte.tv'
d5822b96	27
d5822b96	28	def _real_extract(self, url):
7a249480 PH	29	mobj = re.match(self._VALID_URL, url)
	30	lang = mobj.group('lang')
	31	video_id = mobj.group('id')
	32
8de64cac PH	33	ref_xml_url = url.replace('/videos/', '/do_delegate/videos/')
8de64cac PH	34	ref_xml_url = ref_xml_url.replace('.html', ',view,asPlayerXml.xml')
a4ff6c47 PH	35	ref_xml_doc = self._download_xml(
a4ff6c47 PH	36	ref_xml_url, video_id, note='Downloading metadata')
df50a412	37	config_node = find_xpath_attr(ref_xml_doc, './/video', 'lang', lang)
8de64cac	38	config_xml_url = config_node.attrib['ref']
878d11ec	39	config = self._download_xml(
a4ff6c47	40	config_xml_url, video_id, note='Downloading configuration')
37b6a661	41
878d11ec	42	formats = [{
5a184030	43	'format_id': q.attrib['quality'],
459af434 JMF	44	# The playpath starts at 'mp4:', if we don't manually
	45	# split the url, rtmpdump will incorrectly parse them
	46	'url': q.text.split('mp4:', 1)[0],
	47	'play_path': 'mp4:' + q.text.split('mp4:', 1)[1],
b2799ff9	48	'ext': 'flv',
878d11ec	49	'quality': 2 if q.attrib['quality'] == 'hd' else 1,
b2799ff9	50	} for q in config.findall('./urls/url')]
878d11ec PH	51	self._sort_formats(formats)
	52
	53	title = config.find('.//name').text
	54	thumbnail = config.find('.//firstThumbnailUrl').text
	55	return {
	56	'id': video_id,
	57	'title': title,
	58	'thumbnail': thumbnail,
7a249480	59	'formats': formats,
515bbe4b	60	}
c40f5cf4 JMF	61
	62
	63	class ArteTVPlus7IE(InfoExtractor):
3798eadc	64	IE_NAME = 'arte.tv:+7'
f66ede43	65	_VALID_URL = r'https?://(?:www\.)?arte\.tv/guide/(?P<lang>fr\|de)/(?:(?:sendungen\|emissions)/)?(?P<id>.?)/(?P<name>.?)(\?.*)?'
c40f5cf4	66
69a0c470 JMF	67	@classmethod
	68	def _extract_url_info(cls, url):
	69	mobj = re.match(cls._VALID_URL, url)
c40f5cf4 JMF	70	lang = mobj.group('lang')
	71	# This is not a real id, it can be for example AJT for the news
	72	# http://www.arte.tv/guide/fr/emissions/AJT/arte-journal
	73	video_id = mobj.group('id')
69a0c470	74	return video_id, lang
c40f5cf4	75
69a0c470 JMF	76	def _real_extract(self, url):
69a0c470 JMF	77	video_id, lang = self._extract_url_info(url)
c40f5cf4	78	webpage = self._download_webpage(url, video_id)
69a0c470 JMF	79	return self._extract_from_webpage(webpage, video_id, lang)
	80
	81	def _extract_from_webpage(self, webpage, video_id, lang):
88ce273d	82	json_url = self._html_search_regex(
f2d9e3a3	83	[r'arte_vp_url=["\'](.*?)["\']', r'data-url=["\']([^"]+)["\']'],
393ca8c9 S	84	webpage, 'json vp url', default=None)
	85	if not json_url:
	86	iframe_url = self._html_search_regex(
	87	r'<iframe[^>]+src=(["\'])(?P<url>.+\bjson_url=.+?)\1',
5e39123b	88	webpage, 'iframe url', group='url')
393ca8c9 S	89	json_url = compat_parse_qs(
393ca8c9 S	90	compat_urllib_parse_urlparse(iframe_url).query)['json_url'][0]
56a8ab7d	91	return self._extract_from_json_url(json_url, video_id, lang)
c40f5cf4	92
56a8ab7d	93	def _extract_from_json_url(self, json_url, video_id, lang):
893f8832	94	info = self._download_json(json_url, video_id)
c40f5cf4 JMF	95	player_info = info['videoJsonPlayer']
c40f5cf4 JMF	96
99b67fec PH	97	upload_date_str = player_info.get('shootingDate')
	98	if not upload_date_str:
	99	upload_date_str = player_info.get('VDA', '').split(' ')[0]
	100
74214d35 S	101	title = player_info['VTI'].strip()
	102	subtitle = player_info.get('VSU', '').strip()
	103	if subtitle:
	104	title += ' - %s' % subtitle
	105
c40f5cf4 JMF	106	info_dict = {
c40f5cf4 JMF	107	'id': player_info['VID'],
74214d35	108	'title': title,
c40f5cf4	109	'description': player_info.get('VDE'),
99b67fec	110	'upload_date': unified_strdate(upload_date_str),
c40f5cf4 JMF	111	'thumbnail': player_info.get('programImage') or player_info.get('VTU', {}).get('IUR'),
c40f5cf4 JMF	112	}
aff2f4f4	113	qfunc = qualities(['HQ', 'MQ', 'EQ', 'SQ'])
c40f5cf4	114
aff2f4f4	115	formats = []
d24a2b20	116	for format_id, format_dict in player_info['VSR'].items():
aff2f4f4 PH	117	f = dict(format_dict)
	118	versionCode = f.get('versionCode')
	119
	120	langcode = {
	121	'fr': 'F',
	122	'de': 'A',
	123	}.get(lang, lang)
	124	lang_rexs = [r'VO?%s' % langcode, r'VO?.-ST%s' % langcode]
	125	lang_pref = (
	126	None if versionCode is None else (
	127	10 if any(re.match(r, versionCode) for r in lang_rexs)
	128	else -10))
	129	source_pref = 0
	130	if versionCode is not None:
	131	# The original version with subtitles has lower relevance
	132	if re.match(r'VO-ST(F\|A)', versionCode):
	133	source_pref -= 10
	134	# The version with sourds/mal subtitles has also lower relevance
	135	elif re.match(r'VO?(F\|A)-STM\1', versionCode):
	136	source_pref -= 9
	137	format = {
	138	'format_id': format_id,
	139	'preference': -10 if f.get('videoFormat') == 'M3U8' else None,
	140	'language_preference': lang_pref,
	141	'format_note': '%s, %s' % (f.get('versionCode'), f.get('versionLibelle')),
	142	'width': int_or_none(f.get('width')),
	143	'height': int_or_none(f.get('height')),
	144	'tbr': int_or_none(f.get('bitrate')),
1b7b1d6e	145	'quality': qfunc(f.get('quality')),
aff2f4f4	146	'source_preference': source_pref,
c40f5cf4	147	}
aff2f4f4 PH	148
	149	if f.get('mediaType') == 'rtmp':
	150	format['url'] = f['streamer']
	151	format['play_path'] = 'mp4:' + f['url']
	152	format['ext'] = 'flv'
c40f5cf4	153	else:
aff2f4f4 PH	154	format['url'] = f['url']
	155
	156	formats.append(format)
	157
c06a9f87	158	self._check_formats(formats, video_id)
aff2f4f4	159	self._sort_formats(formats)
c40f5cf4	160
aff2f4f4	161	info_dict['formats'] = formats
c40f5cf4 JMF	162	return info_dict
	163
	164
	165	# It also uses the arte_vp_url url from the webpage to extract the information
	166	class ArteTVCreativeIE(ArteTVPlus7IE):
3798eadc	167	IE_NAME = 'arte.tv:creative'
01d906ff	168	_VALID_URL = r'https?://creative\.arte\.tv/(?P<lang>fr\|de)/(?:magazine?/)?(?P<id>[^?#]+)'
c40f5cf4	169
01d906ff	170	_TESTS = [{
3798eadc	171	'url': 'http://creative.arte.tv/de/magazin/agentur-amateur-corporate-design',
3798eadc	172	'info_dict': {
01d906ff	173	'id': '72176',
39a743fb	174	'ext': 'mp4',
01d906ff PH	175	'title': 'Folge 2 - Corporate Design',
01d906ff PH	176	'upload_date': '20131004',
c40f5cf4	177	},
01d906ff PH	178	}, {
	179	'url': 'http://creative.arte.tv/fr/Monty-Python-Reunion',
	180	'info_dict': {
	181	'id': '160676',
	182	'ext': 'mp4',
	183	'title': 'Monty Python live (mostly)',
b5f4775b PH	184	'description': 'Événement ! Quarante-cinq ans après leurs premiers succès, les légendaires Monty Python remontent sur scène.\n',
b5f4775b PH	185	'upload_date': '20140805',
01d906ff PH	186	}
01d906ff PH	187	}]
c40f5cf4	188
69a0c470 JMF	189
69a0c470 JMF	190	class ArteTVFutureIE(ArteTVPlus7IE):
3798eadc	191	IE_NAME = 'arte.tv:future'
69a0c470 JMF	192	_VALID_URL = r'https?://future\.arte\.tv/(?P<lang>fr\|de)/(thema\|sujet)/.*?#article-anchor-(?P<id>\d+)'
	193
	194	_TEST = {
3798eadc	195	'url': 'http://future.arte.tv/fr/sujet/info-sciences#article-anchor-7081',
3798eadc	196	'info_dict': {
458ade63	197	'id': '5201',
39a743fb	198	'ext': 'mp4',
3798eadc	199	'title': 'Les champignons au secours de la planète',
458ade63	200	'upload_date': '20131101',
69a0c470 JMF	201	},
	202	}
	203
	204	def _real_extract(self, url):
	205	anchor_id, lang = self._extract_url_info(url)
	206	webpage = self._download_webpage(url, anchor_id)
2ad5708c S	207	row = self._search_regex(
	208	r'(?s)id="%s"[^>]>.+?(<div[^>]arte_vp_url[^>]*>)' % anchor_id,
	209	webpage, 'row')
69a0c470	210	return self._extract_from_webpage(row, anchor_id, lang)
56a8ab7d	211
ac5118bc	212
56a8ab7d	213	class ArteTVDDCIE(ArteTVPlus7IE):
3798eadc	214	IE_NAME = 'arte.tv:ddc'
39a743fb	215	_VALID_URL = r'https?://ddc\.arte\.tv/(?P<lang>emission\|folge)/(?P<id>.+)'
56a8ab7d	216
56a8ab7d CD	217	def _real_extract(self, url):
	218	video_id, lang = self._extract_url_info(url)
	219	if lang == 'folge':
	220	lang = 'de'
	221	elif lang == 'emission':
	222	lang = 'fr'
	223	webpage = self._download_webpage(url, video_id)
	224	scriptElement = get_element_by_attribute('class', 'visu_video_block', webpage)
	225	script_url = self._html_search_regex(r'src="(.*?)"', scriptElement, 'script url')
	226	javascriptPlayerGenerator = self._download_webpage(script_url, video_id, 'Download javascript player generator')
	227	json_url = self._search_regex(r"json_url=(.)&rendering_place.", javascriptPlayerGenerator, 'json url')
	228	return self._extract_from_json_url(json_url, video_id, lang)
4966a0b2 JMF	229
	230
	231	class ArteTVConcertIE(ArteTVPlus7IE):
	232	IE_NAME = 'arte.tv:concert'
	233	_VALID_URL = r'https?://concert\.arte\.tv/(?P<lang>de\|fr)/(?P<id>.+)'
	234
	235	_TEST = {
	236	'url': 'http://concert.arte.tv/de/notwist-im-pariser-konzertclub-divan-du-monde',
	237	'md5': '9ea035b7bd69696b67aa2ccaaa218161',
	238	'info_dict': {
	239	'id': '186',
	240	'ext': 'mp4',
	241	'title': 'The Notwist im Pariser Konzertclub "Divan du Monde"',
	242	'upload_date': '20140128',
a9c2896e	243	'description': 'md5:486eb08f991552ade77439fe6d82c305',
4966a0b2 JMF	244	},
4966a0b2 JMF	245	}
893f8832 PH	246
	247
	248	class ArteTVEmbedIE(ArteTVPlus7IE):
	249	IE_NAME = 'arte.tv:embed'
	250	_VALID_URL = r'''(?x)
	251	http://www\.arte\.tv
	252	/playerv2/embed\.php\?json_url=
	253	(?P<json_url>
	254	http://arte\.tv/papi/tvguide/videos/stream/player/
	255	(?P<lang>[^/]+)/(?P<id>[^/]+)[^&]*
	256	)
	257	'''
	258
	259	def _real_extract(self, url):
	260	mobj = re.match(self._VALID_URL, url)
	261	video_id = mobj.group('id')
	262	lang = mobj.group('lang')
	263	json_url = mobj.group('json_url')
	264	return self._extract_from_json_url(json_url, video_id, lang)