diff options
Diffstat (limited to 'youtube_dl/extractor/arte.py')
| -rw-r--r-- | youtube_dl/extractor/arte.py | 415 |
1 files changed, 226 insertions, 189 deletions
diff --git a/youtube_dl/extractor/arte.py b/youtube_dl/extractor/arte.py index 2a00da3ee..5bfe57b10 100644 --- a/youtube_dl/extractor/arte.py +++ b/youtube_dl/extractor/arte.py @@ -1,139 +1,154 @@ -# encoding: utf-8 +# coding: utf-8 from __future__ import unicode_literals import re from .common import InfoExtractor from ..compat import ( - compat_parse_qs, - compat_urllib_parse_urlparse, + compat_str, + compat_urlparse, ) from ..utils import ( - find_xpath_attr, - unified_strdate, - get_element_by_attribute, + ExtractorError, int_or_none, qualities, + strip_or_none, + try_get, + unified_strdate, + url_or_none, ) -# 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 ArteTVBaseIE(InfoExtractor): + _ARTE_LANGUAGES = 'fr|de|en|es|it|pl' + _API_BASE = 'https://api.arte.tv/api/player/v1' -class ArteTvIE(InfoExtractor): - _VALID_URL = r'http://videos\.arte\.tv/(?P<lang>fr|de)/.*-(?P<id>.*?)\.html' - IE_NAME = 'arte.tv' + +class ArteTVIE(ArteTVBaseIE): + _VALID_URL = r'''(?x) + https?:// + (?: + (?:www\.)?arte\.tv/(?P<lang>%(langs)s)/videos| + api\.arte\.tv/api/player/v\d+/config/(?P<lang_2>%(langs)s) + ) + /(?P<id>\d{6}-\d{3}-[AF]) + ''' % {'langs': ArteTVBaseIE._ARTE_LANGUAGES} + _TESTS = [{ + 'url': 'https://www.arte.tv/en/videos/088501-000-A/mexico-stealing-petrol-to-survive/', + 'info_dict': { + 'id': '088501-000-A', + 'ext': 'mp4', + 'title': 'Mexico: Stealing Petrol to Survive', + 'upload_date': '20190628', + }, + }, { + 'url': 'https://www.arte.tv/pl/videos/100103-000-A/usa-dyskryminacja-na-porodowce/', + 'only_matching': True, + }, { + 'url': 'https://api.arte.tv/api/player/v2/config/de/100605-013-A', + 'only_matching': True, + }] 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 + lang = mobj.group('lang') or mobj.group('lang_2') - 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) + info = self._download_json( + '%s/config/%s/%s' % (self._API_BASE, lang, video_id), video_id) player_info = info['videoJsonPlayer'] + vsr = try_get(player_info, lambda x: x['VSR'], dict) + if not vsr: + error = None + if try_get(player_info, lambda x: x['custom_msg']['type']) == 'error': + error = try_get( + player_info, lambda x: x['custom_msg']['msg'], compat_str) + if not error: + error = 'Video %s is not available' % player_info.get('VID') or video_id + raise ExtractorError(error, expected=True) + upload_date_str = player_info.get('shootingDate') if not upload_date_str: - upload_date_str = player_info.get('VDA', '').split(' ')[0] + upload_date_str = (player_info.get('VRA') or player_info.get('VDA') or '').split(' ')[0] - title = player_info['VTI'].strip() + title = (player_info.get('VTI') or player_info['VID']).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(['MQ', 'HQ', 'EQ', 'SQ']) + + LANGS = { + 'fr': 'F', + 'de': 'A', + 'en': 'E[ANG]', + 'es': 'E[ESP]', + 'it': 'E[ITA]', + 'pl': 'E[POL]', } - qfunc = qualities(['HQ', 'MQ', 'EQ', 'SQ']) + + langcode = LANGS.get(lang, lang) formats = [] - for format_id, format_dict in player_info['VSR'].items(): + for format_id, format_dict in vsr.items(): f = dict(format_dict) + format_url = url_or_none(f.get('url')) + streamer = f.get('streamer') + if not format_url and not streamer: + continue versionCode = f.get('versionCode') + l = re.escape(langcode) + + # Language preference from most to least priority + # Reference: section 6.8 of + # https://www.arte.tv/sites/en/corporate/files/complete-technical-guidelines-arte-geie-v1-07-1.pdf + PREFERENCES = ( + # original version in requested language, without subtitles + r'VO{0}$'.format(l), + # original version in requested language, with partial subtitles in requested language + r'VO{0}-ST{0}$'.format(l), + # original version in requested language, with subtitles for the deaf and hard-of-hearing in requested language + r'VO{0}-STM{0}$'.format(l), + # non-original (dubbed) version in requested language, without subtitles + r'V{0}$'.format(l), + # non-original (dubbed) version in requested language, with subtitles partial subtitles in requested language + r'V{0}-ST{0}$'.format(l), + # non-original (dubbed) version in requested language, with subtitles for the deaf and hard-of-hearing in requested language + r'V{0}-STM{0}$'.format(l), + # original version in requested language, with partial subtitles in different language + r'VO{0}-ST(?!{0}).+?$'.format(l), + # original version in requested language, with subtitles for the deaf and hard-of-hearing in different language + r'VO{0}-STM(?!{0}).+?$'.format(l), + # original version in different language, with partial subtitles in requested language + r'VO(?:(?!{0}).+?)?-ST{0}$'.format(l), + # original version in different language, with subtitles for the deaf and hard-of-hearing in requested language + r'VO(?:(?!{0}).+?)?-STM{0}$'.format(l), + # original version in different language, without subtitles + r'VO(?:(?!{0}))?$'.format(l), + # original version in different language, with partial subtitles in different language + r'VO(?:(?!{0}).+?)?-ST(?!{0}).+?$'.format(l), + # original version in different language, with subtitles for the deaf and hard-of-hearing in different language + r'VO(?:(?!{0}).+?)?-STM(?!{0}).+?$'.format(l), + ) + + for pref, p in enumerate(PREFERENCES): + if re.match(p, versionCode): + lang_pref = len(PREFERENCES) - pref + break + else: + lang_pref = -1 + + media_type = f.get('mediaType') + if media_type == 'hls': + m3u8_formats = self._extract_m3u8_formats( + format_url, video_id, 'mp4', entry_protocol='m3u8_native', + m3u8_id=format_id, fatal=False) + for m3u8_format in m3u8_formats: + m3u8_format['language_preference'] = lang_pref + formats.extend(m3u8_formats) + continue - 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, @@ -143,10 +158,9 @@ class ArteTVPlus7IE(InfoExtractor): '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': + if media_type == 'rtmp': format['url'] = f['streamer'] format['play_path'] = 'mp4:' + f['url'] format['ext'] = 'flv' @@ -155,110 +169,133 @@ class ArteTVPlus7IE(InfoExtractor): formats.append(format) - self._check_formats(formats, video_id) self._sort_formats(formats) - info_dict['formats'] = formats - return info_dict - + return { + 'id': player_info.get('VID') or video_id, + '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'), + 'formats': formats, + } -# 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>[^?#]+)' +class ArteTVEmbedIE(InfoExtractor): + _VALID_URL = r'https?://(?:www\.)?arte\.tv/player/v\d+/index\.php\?.*?\bjson_url=.+' _TESTS = [{ - 'url': 'http://creative.arte.tv/de/magazin/agentur-amateur-corporate-design', + 'url': 'https://www.arte.tv/player/v5/index.php?json_url=https%3A%2F%2Fapi.arte.tv%2Fapi%2Fplayer%2Fv2%2Fconfig%2Fde%2F100605-013-A&lang=de&autoplay=true&mute=0100605-013-A', 'info_dict': { - 'id': '72176', + 'id': '100605-013-A', 'ext': 'mp4', - 'title': 'Folge 2 - Corporate Design', - 'upload_date': '20131004', + 'title': 'United we Stream November Lockdown Edition #13', + 'description': 'md5:be40b667f45189632b78c1425c7c2ce1', + 'upload_date': '20201116', }, }, { - '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', - } + 'url': 'https://www.arte.tv/player/v3/index.php?json_url=https://api.arte.tv/api/player/v2/config/de/100605-013-A', + 'only_matching': True, }] - -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', - }, - } + @staticmethod + def _extract_urls(webpage): + return [url for _, url in re.findall( + r'<(?:iframe|script)[^>]+src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?arte\.tv/player/v\d+/index\.php\?.*?\bjson_url=.+?)\1', + webpage)] 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) + qs = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query) + json_url = qs['json_url'][0] + video_id = ArteTVIE._match_id(json_url) + return self.url_result( + json_url, ie=ArteTVIE.ie_key(), video_id=video_id) -class ArteTVDDCIE(ArteTVPlus7IE): - IE_NAME = 'arte.tv:ddc' - _VALID_URL = r'https?://ddc\.arte\.tv/(?P<lang>emission|folge)/(?P<id>.+)' +class ArteTVPlaylistIE(ArteTVBaseIE): + _VALID_URL = r'https?://(?:www\.)?arte\.tv/(?P<lang>%s)/videos/(?P<id>RC-\d{6})' % ArteTVBaseIE._ARTE_LANGUAGES + _TESTS = [{ + 'url': 'https://www.arte.tv/en/videos/RC-016954/earn-a-living/', + 'info_dict': { + 'id': 'RC-016954', + 'title': 'Earn a Living', + 'description': 'md5:d322c55011514b3a7241f7fb80d494c2', + }, + 'playlist_mincount': 6, + }, { + 'url': 'https://www.arte.tv/pl/videos/RC-014123/arte-reportage/', + 'only_matching': True, + }] 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', + lang, playlist_id = re.match(self._VALID_URL, url).groups() + collection = self._download_json( + '%s/collectionData/%s/%s?source=videos' + % (self._API_BASE, lang, playlist_id), playlist_id) + entries = [] + for video in collection['videos']: + if not isinstance(video, dict): + continue + video_url = url_or_none(video.get('url')) or url_or_none(video.get('jsonUrl')) + if not video_url: + continue + video_id = video.get('programId') + entries.append({ + '_type': 'url_transparent', + 'url': video_url, + 'id': video_id, + 'title': video.get('title'), + 'alt_title': video.get('subtitle'), + 'thumbnail': url_or_none(try_get(video, lambda x: x['mainImage']['url'], compat_str)), + 'duration': int_or_none(video.get('durationSeconds')), + 'view_count': int_or_none(video.get('views')), + 'ie_key': ArteTVIE.ie_key(), + }) + title = collection.get('title') + description = collection.get('shortDescription') or collection.get('teaserText') + return self.playlist_result(entries, playlist_id, title, description) + + +class ArteTVCategoryIE(ArteTVBaseIE): + _VALID_URL = r'https?://(?:www\.)?arte\.tv/(?P<lang>%s)/videos/(?P<id>[\w-]+(?:/[\w-]+)*)/?\s*$' % ArteTVBaseIE._ARTE_LANGUAGES + _TESTS = [{ + 'url': 'https://www.arte.tv/en/videos/politics-and-society/', 'info_dict': { - 'id': '186', - 'ext': 'mp4', - 'title': 'The Notwist im Pariser Konzertclub "Divan du Monde"', - 'upload_date': '20140128', - 'description': 'md5:486eb08f991552ade77439fe6d82c305', + 'id': 'politics-and-society', + 'title': 'Politics and society', + 'description': 'Investigative documentary series, geopolitical analysis, and international commentary', }, - } + 'playlist_mincount': 13, + }, + ] - -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>[^/]+)[^&]* - ) - ''' + @classmethod + def suitable(cls, url): + return ( + not any(ie.suitable(url) for ie in (ArteTVIE, ArteTVPlaylistIE, )) + and super(ArteTVCategoryIE, cls).suitable(url)) 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) + lang, playlist_id = re.match(self._VALID_URL, url).groups() + webpage = self._download_webpage(url, playlist_id) + + items = [] + for video in re.finditer( + r'<a\b[^>]*?href\s*=\s*(?P<q>"|\'|\b)(?P<url>https?://www\.arte\.tv/%s/videos/[\w/-]+)(?P=q)' % lang, + webpage): + video = video.group('url') + if video == url: + continue + if any(ie.suitable(video) for ie in (ArteTVIE, ArteTVPlaylistIE, )): + items.append(video) + + if items: + title = (self._og_search_title(webpage, default=None) + or self._html_search_regex(r'<title\b[^>]*>([^<]+)</title>', default=None)) + title = strip_or_none(title.rsplit('|', 1)[0]) or self._generic_title(url) + + result = self.playlist_from_matches(items, playlist_id=playlist_id, playlist_title=title) + if result: + description = self._og_search_description(webpage, default=None) + if description: + result['description'] = description + return result |