X-Git-Url: http://git.bitcoin.ninja/index.cgi?a=blobdiff_plain;f=youtube_dl%2Fextractor%2Farte.py;h=2bd3bfe8a852159cf7aefa93e731bab995ca7d23;hb=HEAD;hp=f4d033e7094ff1614cda4043a74c37dd7b6d16ad;hpb=b72f6163dc264218c4abefebefd1971875fff7d9;p=youtube-dl diff --git a/youtube_dl/extractor/arte.py b/youtube_dl/extractor/arte.py index f4d033e70..2bd3bfe8a 100644 --- a/youtube_dl/extractor/arte.py +++ b/youtube_dl/extractor/arte.py @@ -1,20 +1,16 @@ -# 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, -) +from ..compat import compat_str from ..utils import ( - find_xpath_attr, - unified_strdate, - get_element_by_attribute, + ExtractorError, int_or_none, - NO_DEFAULT, qualities, + try_get, + unified_strdate, ) # There are different sources of video in arte.tv, the extraction process @@ -22,113 +18,26 @@ from ..utils import ( # add tests. -class ArteTvIE(InfoExtractor): - _VALID_URL = r'http://videos\.arte\.tv/(?Pfr|de|en|es)/.*-(?P.*?)\.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/(?Pfr|de|en|es)/(?:(?:sendungen|emissions|embed)/)?(?P[^/]+)/(?P[^/?#&+])' - - @classmethod - def _extract_url_info(cls, url): - mobj = re.match(cls._VALID_URL, url) - lang = mobj.group('lang') - query = compat_parse_qs(compat_urllib_parse_urlparse(url).query) - if 'vid' in query: - video_id = query['vid'][0] - else: - # 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): - patterns_templates = (r'arte_vp_url=["\'](.*?%s.*?)["\']', r'data-url=["\']([^"]+%s[^"]+)["\']') - ids = (video_id, '') - # some pages contain multiple videos (like - # http://www.arte.tv/guide/de/sendungen/XEN/xenius/?vid=055918-015_PLUS7-D), - # so we first try to look for json URLs that contain the video id from - # the 'vid' parameter. - patterns = [t % re.escape(_id) for _id in ids for t in patterns_templates] - json_url = self._html_search_regex( - patterns, webpage, 'json vp url', default=None) - if not json_url: - def find_iframe_url(webpage, default=NO_DEFAULT): - return self._html_search_regex( - r']+src=(["\'])(?P.+\bjson_url=.+?)\1', - webpage, 'iframe url', group='url', default=default) - - iframe_url = find_iframe_url(webpage, None) - if not iframe_url: - embed_url = self._html_search_regex( - r'arte_vp_url_oembed=\'([^\']+?)\'', webpage, 'embed url', default=None) - if embed_url: - player = self._download_json( - embed_url, video_id, 'Downloading player page') - iframe_url = find_iframe_url(player['html']) - # en and es URLs produce react-based pages with different layout (e.g. - # http://www.arte.tv/guide/en/053330-002-A/carnival-italy?zone=world) - if not iframe_url: - embed_html = self._parse_json( - self._search_regex( - r'program\s*:\s*({.+?["\']embed_html["\'].+?}),?\s*\n', - webpage, 'program'), - video_id)['embed_html'] - iframe_url = find_iframe_url(embed_html) - 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): +class ArteTVBaseIE(InfoExtractor): + def _extract_from_json_url(self, json_url, video_id, lang, title=None): info = self._download_json(json_url, 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('VRA') or player_info.get('VDA') or '').split(' ')[0] - title = player_info['VTI'].strip() + title = (player_info.get('VTI') or title or player_info['VID']).strip() subtitle = player_info.get('VSU', '').strip() if subtitle: title += ' - %s' % subtitle @@ -140,33 +49,64 @@ class ArteTVPlus7IE(InfoExtractor): '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']) + qfunc = qualities(['MQ', 'HQ', 'EQ', 'SQ']) LANGS = { 'fr': 'F', 'de': 'A', 'en': 'E[ANG]', 'es': 'E[ESP]', + 'it': 'E[ITA]', + 'pl': 'E[POL]', } + 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) versionCode = f.get('versionCode') - langcode = LANGS.get(lang, lang) - lang_rexs = [r'VO?%s-' % re.escape(langcode), r'VO?.-ST%s$' % re.escape(langcode)] - lang_pref = None - if versionCode: - matched_lang_rexs = [r for r in lang_rexs if re.match(r, versionCode)] - lang_pref = -10 if not matched_lang_rexs else 10 * len(matched_lang_rexs) - source_pref = 0 - if versionCode is not None: - # The original version with subtitles has lower relevance - if re.match(r'VO-ST(F|A|E)', versionCode): - source_pref -= 10 - # The version with sourds/mal subtitles has also lower relevance - elif re.match(r'VO?(F|A|E)-STM\1', versionCode): - source_pref -= 9 + l = re.escape(langcode) + + # Language preference from most to least priority + # Reference: section 6.8 of + # https://www.arte.tv/sites/en/corporate/files/complete-technical-guidelines-arte-geie-v1-07-1.pdf + PREFERENCES = ( + # original version in requested language, without subtitles + r'VO{0}$'.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 + format = { 'format_id': format_id, 'preference': -10 if f.get('videoFormat') == 'M3U8' else None, @@ -176,7 +116,6 @@ 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': @@ -195,132 +134,68 @@ class ArteTVPlus7IE(InfoExtractor): 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/(?Pfr|de|en|es)/(?:magazine?/)?(?P[^/?#&]+)' +class ArteTVPlus7IE(ArteTVBaseIE): + IE_NAME = 'arte.tv:+7' + _VALID_URL = r'https?://(?:www\.)?arte\.tv/(?Pfr|de|en|es|it|pl)/videos/(?P\d{6}-\d{3}-[AF])' _TESTS = [{ - 'url': 'http://creative.arte.tv/de/magazin/agentur-amateur-corporate-design', + 'url': 'https://www.arte.tv/en/videos/088501-000-A/mexico-stealing-petrol-to-survive/', 'info_dict': { - 'id': '72176', + 'id': '088501-000-A', 'ext': 'mp4', - 'title': 'Folge 2 - Corporate Design', - 'upload_date': '20131004', + 'title': 'Mexico: Stealing Petrol to Survive', + 'upload_date': '20190628', }, - }, { - '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', - } }] + def _real_extract(self, url): + lang, video_id = re.match(self._VALID_URL, url).groups() + return self._extract_from_json_url( + 'https://api.arte.tv/api/player/v1/config/%s/%s' % (lang, video_id), + video_id, lang) -class ArteTVFutureIE(ArteTVPlus7IE): - IE_NAME = 'arte.tv:future' - _VALID_URL = r'https?://future\.arte\.tv/(?Pfr|de|en|es)/(?P[^/?#&]+)' - - _TESTS = [{ - 'url': 'http://future.arte.tv/fr/info-sciences/les-ecrevisses-aussi-sont-anxieuses', - 'info_dict': { - 'id': '050940-028-A', - 'ext': 'mp4', - 'title': 'Les écrevisses aussi peuvent être anxieuses', - }, - }, { - 'url': 'http://future.arte.tv/fr/la-science-est-elle-responsable', - 'only_matching': True, - }] +class ArteTVEmbedIE(ArteTVPlus7IE): + IE_NAME = 'arte.tv:embed' + _VALID_URL = r'''(?x) + https://www\.arte\.tv + /player/v3/index\.php\?json_url= + (?P + https?://api\.arte\.tv/api/player/v1/config/ + (?P[^/]+)/(?P\d{6}-\d{3}-[AF]) + ) + ''' -class ArteTVDDCIE(ArteTVPlus7IE): - IE_NAME = 'arte.tv:ddc' - _VALID_URL = r'https?://ddc\.arte\.tv/(?Pemission|folge)/(?P[^/?#&]+)' + _TESTS = [] 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') + json_url, lang, video_id = re.match(self._VALID_URL, url).groups() 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/(?Pfr|de|en|es)/(?P[^/?#&]+)' - - _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 ArteTVCinemaIE(ArteTVPlus7IE): - IE_NAME = 'arte.tv:cinema' - _VALID_URL = r'https?://cinema\.arte\.tv/(?Pfr|de|en|es)/(?P.+)' - - _TEST = { - 'url': 'http://cinema.arte.tv/de/node/38291', - 'md5': '6b275511a5107c60bacbeeda368c3aa1', - 'info_dict': { - 'id': '055876-000_PWA12025-D', - 'ext': 'mp4', - 'title': 'Tod auf dem Nil', - 'upload_date': '20160122', - 'description': 'md5:7f749bbb77d800ef2be11d54529b96bc', - }, - } - - -class ArteTVMagazineIE(ArteTVPlus7IE): - IE_NAME = 'arte.tv:magazine' - _VALID_URL = r'https?://(?:www\.)?arte\.tv/magazine/[^/]+/(?Pfr|de|en|es)/(?P[^/?#&]+)' +class ArteTVPlaylistIE(ArteTVBaseIE): + IE_NAME = 'arte.tv:playlist' + _VALID_URL = r'https?://(?:www\.)?arte\.tv/(?Pfr|de|en|es|it|pl)/videos/(?PRC-\d{6})' _TESTS = [{ - 'url': 'http://www.arte.tv/magazine/trepalium/fr/entretien-avec-le-realisateur-vincent-lannoo-trepalium', - 'md5': '66a093339c1278bb3719157ef07107b2', + 'url': 'https://www.arte.tv/en/videos/RC-016954/earn-a-living/', 'info_dict': { - 'id': '065965-000-A', - 'ext': 'mp4', - 'title': 'Trepalium - Extrait Ep.01', + 'id': 'RC-016954', + 'title': 'Earn a Living', + 'description': 'md5:d322c55011514b3a7241f7fb80d494c2', }, - }, { - 'url': 'http://www.arte.tv/magazine/metropolis/de/frank-woeste-german-paris-metropolis', - 'only_matching': True, + 'playlist_mincount': 6, }] - -class ArteTVEmbedIE(ArteTVPlus7IE): - IE_NAME = 'arte.tv:embed' - _VALID_URL = r'''(?x) - http://www\.arte\.tv - /playerv2/embed\.php\?json_url= - (?P - http://arte\.tv/papi/tvguide/videos/stream/player/ - (?P[^/]+)/(?P[^/]+)[^&]* - ) - ''' - 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() + collection = self._download_json( + 'https://api.arte.tv/api/player/v1/collectionData/%s/%s?source=videos' + % (lang, playlist_id), playlist_id) + title = collection.get('title') + description = collection.get('shortDescription') or collection.get('teaserText') + entries = [ + self._extract_from_json_url( + video['jsonUrl'], video.get('programId') or playlist_id, lang) + for video in collection['videos'] if video.get('jsonUrl')] + return self.playlist_result(entries, playlist_id, title, description)