X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/975f9519f0ae48e780767bf543204c6626a493b0..848723ea972c09f28787db91d8c06e98a274ab89:/youtube_dl/extractor/arte.py?ds=sidebyside diff --git a/youtube_dl/extractor/arte.py b/youtube_dl/extractor/arte.py index e7a91a1..7cf3785 100644 --- a/youtube_dl/extractor/arte.py +++ b/youtube_dl/extractor/arte.py @@ -1,28 +1,34 @@ +# encoding: utf-8 +from __future__ import unicode_literals + import re import json -import xml.etree.ElementTree from .common import InfoExtractor from ..utils import ( ExtractorError, + find_xpath_attr, unified_strdate, + determine_ext, + get_element_by_id, + compat_str, + get_element_by_attribute, ) +# 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): - """ - There are two sources of video in arte.tv: videos.arte.tv and - www.arte.tv/guide, the extraction process is different for each one. - The videos expire in 7 days, so we can't add tests. - """ - _EMISSION_URL = r'(?:http://)?www\.arte.tv/guide/(?Pfr|de)/(?:(?:sendungen|emissions)/)?(?P.*?)/(?P.*?)(\?.*)?' - _VIDEOS_URL = r'(?:http://)?videos.arte.tv/(?Pfr|de)/.*-(?P.*?).html' + _VIDEOS_URL = r'(?:http://)?videos\.arte\.tv/(?Pfr|de)/.*-(?P.*?)\.html' + _LIVEWEB_URL = r'(?:http://)?liveweb\.arte\.tv/(?Pfr|de)/(?P.+?)/(?P.+)' _LIVE_URL = r'index-[0-9]+\.html$' - IE_NAME = u'arte.tv' + IE_NAME = 'arte.tv' @classmethod def suitable(cls, url): - return any(re.match(regex, url) for regex in (cls._EMISSION_URL, cls._VIDEOS_URL)) + return any(re.match(regex, url) for regex in (cls._VIDEOS_URL, cls._LIVEWEB_URL)) # TODO implement Live Stream # from ..utils import compat_urllib_parse @@ -33,7 +39,7 @@ class ArteTvIE(InfoExtractor): # r'src="(.*?/videothek_js.*?\.js)', # 0, # [ - # (1, 'url', u'Invalid URL: %s' % url) + # (1, 'url', 'Invalid URL: %s' % url) # ] # ) # http_host = url.split('/')[2] @@ -45,81 +51,37 @@ class ArteTvIE(InfoExtractor): # '(rtmp://.*?)\'', # re.DOTALL, # [ - # (1, 'path', u'could not extract video path: %s' % url), - # (2, 'player', u'could not extract video player: %s' % url), - # (3, 'url', u'could not extract video url: %s' % url) + # (1, 'path', 'could not extract video path: %s' % url), + # (2, 'player', 'could not extract video player: %s' % url), + # (3, 'url', 'could not extract video url: %s' % url) # ] # ) - # video_url = u'%s/%s' % (info.get('url'), info.get('path')) + # video_url = '%s/%s' % (info.get('url'), info.get('path')) def _real_extract(self, url): - mobj = re.match(self._EMISSION_URL, url) - if mobj is not None: - 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 self._extract_emission(url, video_id, lang) - mobj = re.match(self._VIDEOS_URL, url) if mobj is not None: id = mobj.group('id') lang = mobj.group('lang') return self._extract_video(url, id, lang) - if re.search(self._LIVE_URL, video_id) is not None: + mobj = re.match(self._LIVEWEB_URL, url) + if mobj is not None: + name = mobj.group('name') + lang = mobj.group('lang') + return self._extract_liveweb(url, name, lang) + + if re.search(self._LIVE_URL, url) is not None: raise ExtractorError(u'Arte live streams are not yet supported, sorry') # self.extractLiveStream(url) # return - def _extract_emission(self, url, video_id, lang): - """Extract from www.arte.tv/guide""" - webpage = self._download_webpage(url, video_id) - json_url = self._html_search_regex(r'arte_vp_url="(.*?)"', webpage, 'json url') - - json_info = self._download_webpage(json_url, video_id, 'Downloading info json') - self.report_extraction(video_id) - info = json.loads(json_info) - player_info = info['videoJsonPlayer'] - - info_dict = {'id': player_info['VID'], - 'title': player_info['VTI'], - 'description': player_info['VDE'], - 'upload_date': unified_strdate(player_info['VDA'].split(' ')[0]), - 'thumbnail': player_info['programImage'], - 'ext': 'flv', - } - - formats = player_info['VSR'].values() - def _match_lang(f): - # Return true if that format is in the language of the url - if lang == 'fr': - l = 'F' - elif lang == 'de': - l = 'A' - regexes = [r'VO?%s' % l, r'V%s-ST.' % l] - return any(re.match(r, f['versionCode']) for r in regexes) - # Some formats may not be in the same language as the url - formats = filter(_match_lang, formats) - # We order the formats by quality - formats = sorted(formats, key=lambda f: int(f['height'])) - # Pick the best quality - format_info = formats[-1] - if format_info['mediaType'] == u'rtmp': - info_dict['url'] = format_info['streamer'] - info_dict['play_path'] = 'mp4:' + format_info['url'] - else: - info_dict['url'] = format_info['url'] - - return info_dict - def _extract_video(self, url, video_id, lang): """Extract from videos.arte.tv""" ref_xml_url = url.replace('/videos/', '/do_delegate/videos/') ref_xml_url = ref_xml_url.replace('.html', ',view,asPlayerXml.xml') - ref_xml = self._download_webpage(ref_xml_url, video_id, note=u'Downloading metadata') - ref_xml_doc = xml.etree.ElementTree.fromstring(ref_xml) - config_node = ref_xml_doc.find('.//video[@lang="%s"]' % lang) + ref_xml_doc = self._download_xml(ref_xml_url, video_id, note=u'Downloading metadata') + config_node = find_xpath_attr(ref_xml_doc, './/video', 'lang', lang) config_xml_url = config_node.attrib['ref'] config_xml = self._download_webpage(config_xml_url, video_id, note=u'Downloading configuration') @@ -143,3 +105,178 @@ class ArteTvIE(InfoExtractor): 'url': video_url, 'ext': 'flv', } + + def _extract_liveweb(self, url, name, lang): + """Extract form http://liveweb.arte.tv/""" + webpage = self._download_webpage(url, name) + video_id = self._search_regex(r'eventId=(\d+?)("|&)', webpage, 'event id') + config_doc = self._download_xml('http://download.liveweb.arte.tv/o21/liveweb/events/event-%s.xml' % video_id, + video_id, 'Downloading information') + event_doc = config_doc.find('event') + url_node = event_doc.find('video').find('urlHd') + if url_node is None: + url_node = event_doc.find('urlSd') + + return {'id': video_id, + 'title': event_doc.find('name%s' % lang.capitalize()).text, + 'url': url_node.text.replace('MP4', 'mp4'), + 'ext': 'flv', + 'thumbnail': self._og_search_thumbnail(webpage), + } + + +class ArteTVPlus7IE(InfoExtractor): + IE_NAME = 'arte.tv:+7' + _VALID_URL = r'https?://www\.arte.tv/guide/(?Pfr|de)/(?:(?:sendungen|emissions)/)?(?P.*?)/(?P.*?)(\?.*)?' + + @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="(.*?)"', webpage, 'json url') + return self._extract_from_json_url(json_url, video_id, lang) + + def _extract_from_json_url(self, json_url, video_id, lang): + json_info = self._download_webpage(json_url, video_id, 'Downloading info json') + self.report_extraction(video_id) + info = json.loads(json_info) + player_info = info['videoJsonPlayer'] + + info_dict = { + 'id': player_info['VID'], + 'title': player_info['VTI'], + 'description': player_info.get('VDE'), + 'upload_date': unified_strdate(player_info.get('VDA', '').split(' ')[0]), + 'thumbnail': player_info.get('programImage') or player_info.get('VTU', {}).get('IUR'), + } + + all_formats = player_info['VSR'].values() + # Some formats use the m3u8 protocol + all_formats = list(filter(lambda f: f.get('videoFormat') != 'M3U8', all_formats)) + def _match_lang(f): + if f.get('versionCode') is None: + return True + # Return true if that format is in the language of the url + if lang == 'fr': + l = 'F' + elif lang == 'de': + l = 'A' + regexes = [r'VO?%s' % l, r'VO?.-ST%s' % l] + return any(re.match(r, f['versionCode']) for r in regexes) + # Some formats may not be in the same language as the url + formats = filter(_match_lang, all_formats) + formats = list(formats) # in python3 filter returns an iterator + if not formats: + # Some videos are only available in the 'Originalversion' + # they aren't tagged as being in French or German + if all(f['versionCode'] == 'VO' for f in all_formats): + formats = all_formats + else: + raise ExtractorError(u'The formats list is empty') + + if re.match(r'[A-Z]Q', formats[0]['quality']) is not None: + def sort_key(f): + return ['HQ', 'MQ', 'EQ', 'SQ'].index(f['quality']) + else: + def sort_key(f): + return ( + # Sort first by quality + int(f.get('height',-1)), + int(f.get('bitrate',-1)), + # The original version with subtitles has lower relevance + re.match(r'VO-ST(F|A)', f.get('versionCode', '')) is None, + # The version with sourds/mal subtitles has also lower relevance + re.match(r'VO?(F|A)-STM\1', f.get('versionCode', '')) is None, + ) + formats = sorted(formats, key=sort_key) + def _format(format_info): + quality = '' + height = format_info.get('height') + if height is not None: + quality = compat_str(height) + bitrate = format_info.get('bitrate') + if bitrate is not None: + quality += '-%d' % bitrate + if format_info.get('versionCode') is not None: + format_id = '%s-%s' % (quality, format_info['versionCode']) + else: + format_id = quality + info = { + 'format_id': format_id, + 'format_note': format_info.get('versionLibelle'), + 'width': format_info.get('width'), + 'height': height, + } + if format_info['mediaType'] == 'rtmp': + info['url'] = format_info['streamer'] + info['play_path'] = 'mp4:' + format_info['url'] + info['ext'] = 'flv' + else: + info['url'] = format_info['url'] + info['ext'] = determine_ext(info['url']) + return info + info_dict['formats'] = [_format(f) for f in 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/(?Pfr|de)/magazine?/(?P.+)' + + _TEST = { + 'url': 'http://creative.arte.tv/de/magazin/agentur-amateur-corporate-design', + 'file': '050489-002.mp4', + 'info_dict': { + 'title': 'Agentur Amateur / Agence Amateur #2 : Corporate Design', + }, + } + + +class ArteTVFutureIE(ArteTVPlus7IE): + IE_NAME = 'arte.tv:future' + _VALID_URL = r'https?://future\.arte\.tv/(?Pfr|de)/(thema|sujet)/.*?#article-anchor-(?P\d+)' + + _TEST = { + 'url': 'http://future.arte.tv/fr/sujet/info-sciences#article-anchor-7081', + 'file': '050940-003.mp4', + 'info_dict': { + 'title': 'Les champignons au secours de la planète', + }, + } + + def _real_extract(self, url): + anchor_id, lang = self._extract_url_info(url) + webpage = self._download_webpage(url, anchor_id) + row = get_element_by_id(anchor_id, webpage) + return self._extract_from_webpage(row, anchor_id, lang) + + +class ArteTVDDCIE(ArteTVPlus7IE): + IE_NAME = 'arte.tv:ddc' + _VALID_URL = r'http?://ddc\.arte\.tv/(?Pemission|folge)/(?P.+)' + + 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)