]> Raphaël G. Git Repositories - youtubedl/blobdiff - youtube_dl/extractor/arte.py
Merge tag 'upstream/2016.02.22'
[youtubedl] / youtube_dl / extractor / arte.py
index 18d5916589b239c9cf12a7caa629efeea784ce86..efde7e207bc8d166e80f2a26429797684535d114 100644 (file)
+# encoding: utf-8
+from __future__ import unicode_literals
+
 import re
-import json
-import xml.etree.ElementTree
 
 from .common import InfoExtractor
+from ..compat import (
+    compat_parse_qs,
+    compat_urllib_parse_urlparse,
+)
 from ..utils import (
-    ExtractorError,
     find_xpath_attr,
     unified_strdate,
+    get_element_by_attribute,
+    int_or_none,
+    NO_DEFAULT,
+    qualities,
 )
 
-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/(?P<lang>fr|de)/(?:(?:sendungen|emissions)/)?(?P<id>.*?)/(?P<name>.*?)(\?.*)?'
-    _VIDEOS_URL = r'(?:http://)?videos.arte.tv/(?P<lang>fr|de)/.*-(?P<id>.*?).html'
-    _LIVE_URL = r'index-[0-9]+\.html$'
+# 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.
 
-    IE_NAME = u'arte.tv'
 
-    @classmethod
-    def suitable(cls, url):
-        return any(re.match(regex, url) for regex in (cls._EMISSION_URL, cls._VIDEOS_URL))
-
-    # TODO implement Live Stream
-    # from ..utils import compat_urllib_parse
-    # def extractLiveStream(self, url):
-    #     video_lang = url.split('/')[-4]
-    #     info = self.grep_webpage(
-    #         url,
-    #         r'src="(.*?/videothek_js.*?\.js)',
-    #         0,
-    #         [
-    #             (1, 'url', u'Invalid URL: %s' % url)
-    #         ]
-    #     )
-    #     http_host = url.split('/')[2]
-    #     next_url = 'http://%s%s' % (http_host, compat_urllib_parse.unquote(info.get('url')))
-    #     info = self.grep_webpage(
-    #         next_url,
-    #         r'(s_artestras_scst_geoFRDE_' + video_lang + '.*?)\'.*?' +
-    #             '(http://.*?\.swf).*?' +
-    #             '(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)
-    #         ]
-    #     )
-    #     video_url = u'%s/%s' % (info.get('url'), info.get('path'))
+class ArteTvIE(InfoExtractor):
+    _VALID_URL = r'http://videos\.arte\.tv/(?P<lang>fr|de|en|es)/.*-(?P<id>.*?)\.html'
+    IE_NAME = 'arte.tv'
 
     def _real_extract(self, url):
-        mobj = re.match(self._EMISSION_URL, url)
-        if mobj is not None:
-            lang = mobj.group('lang')
+        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|en|es)/(?:(?:sendungen|emissions|embed)/)?(?P<id>[^/]+)/(?P<name>[^/?#&+])'
+
+    @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 self._extract_emission(url, video_id, lang)
+        return 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)
+    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)
 
-        if re.search(self._LIVE_URL, video_id) is not None:
-            raise ExtractorError(u'Arte live streams are not yet supported, sorry')
-            # self.extractLiveStream(url)
-            # return
+    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'<iframe[^>]+src=(["\'])(?P<url>.+\bjson_url=.+?)\1',
+                    webpage, 'iframe url', group='url', default=default)
 
-    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')
+            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:
+                program = self._search_regex(
+                    r'program\s*:\s*({.+?["\']embed_html["\'].+?}),?\s*\n',
+                    webpage, 'program', default=None)
+                if program:
+                    embed_html = self._parse_json(program, video_id)
+                    if embed_html:
+                        iframe_url = find_iframe_url(embed_html['embed_html'])
+            if iframe_url:
+                json_url = compat_parse_qs(
+                    compat_urllib_parse_urlparse(iframe_url).query)['json_url'][0]
+        if json_url:
+            return self._extract_from_json_url(json_url, video_id, lang)
+        # Differend kind of embed URL (e.g.
+        # http://www.arte.tv/magazine/trepalium/fr/episode-0406-replay-trepalium)
+        embed_url = self._search_regex(
+            r'<iframe[^>]+src=(["\'])(?P<url>.+?)\1',
+            webpage, 'embed url', group='url')
+        return self.url_result(embed_url)
 
-        json_info = self._download_webpage(json_url, video_id, 'Downloading info json')
-        self.report_extraction(video_id)
-        info = json.loads(json_info)
+    def _extract_from_json_url(self, json_url, video_id, lang):
+        info = self._download_json(json_url, video_id)
         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'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, 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']
+        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]
 
-        return info_dict
+        title = player_info['VTI'].strip()
+        subtitle = player_info.get('VSU', '').strip()
+        if subtitle:
+            title += ' - %s' % subtitle
 
-    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 = 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')
+        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'])
+
+        LANGS = {
+            'fr': 'F',
+            'de': 'A',
+            'en': 'E[ANG]',
+            'es': 'E[ESP]',
+        }
+
+        formats = []
+        for format_id, format_dict in player_info['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
+            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,
+            }
 
-        video_urls = list(re.finditer(r'<url quality="(?P<quality>.*?)">(?P<url>.*?)</url>', config_xml))
-        def _key(m):
-            quality = m.group('quality')
-            if quality == 'hd':
-                return 2
+            if f.get('mediaType') == 'rtmp':
+                format['url'] = f['streamer']
+                format['play_path'] = 'mp4:' + f['url']
+                format['ext'] = 'flv'
             else:
-                return 1
-        # We pick the best quality
-        video_urls = sorted(video_urls, key=_key)
-        video_url = list(video_urls)[-1].group('url')
-        
-        title = self._html_search_regex(r'<name>(.*?)</name>', config_xml, 'title')
-        thumbnail = self._html_search_regex(r'<firstThumbnailUrl>(.*?)</firstThumbnailUrl>',
-                                            config_xml, 'thumbnail')
-        return {'id': video_id,
-                'title': title,
-                'thumbnail': thumbnail,
-                'url': video_url,
-                'ext': 'flv',
-                }
+                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|en|es)/(?: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|en|es)/(?P<id>[^/?#&]+)'
+
+    _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',
+            'upload_date': '20140902',
+        },
+    }, {
+        'url': 'http://future.arte.tv/fr/la-science-est-elle-responsable',
+        'only_matching': True,
+    }]
+
+
+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>fr|de|en|es)/(?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 ArteTVCinemaIE(ArteTVPlus7IE):
+    IE_NAME = 'arte.tv:cinema'
+    _VALID_URL = r'https?://cinema\.arte\.tv/(?P<lang>fr|de|en|es)/(?P<id>.+)'
+
+    _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/[^/]+/(?P<lang>fr|de|en|es)/(?P<id>[^/?#&]+)'
+
+    _TESTS = [{
+        # Embedded via <iframe src="http://www.arte.tv/arte_vp/index.php?json_url=..."
+        'url': 'http://www.arte.tv/magazine/trepalium/fr/entretien-avec-le-realisateur-vincent-lannoo-trepalium',
+        'md5': '2a9369bcccf847d1c741e51416299f25',
+        'info_dict': {
+            'id': '065965-000-A',
+            'ext': 'mp4',
+            'title': 'Trepalium - Extrait Ep.01',
+            'upload_date': '20160121',
+        },
+    }, {
+        # Embedded via <iframe src="http://www.arte.tv/guide/fr/embed/054813-004-A/medium"
+        'url': 'http://www.arte.tv/magazine/trepalium/fr/episode-0406-replay-trepalium',
+        'md5': 'fedc64fc7a946110fe311634e79782ca',
+        'info_dict': {
+            'id': '054813-004_PLUS7-F',
+            'ext': 'mp4',
+            'title': 'Trepalium (4/6)',
+            'description': 'md5:10057003c34d54e95350be4f9b05cb40',
+            'upload_date': '20160218',
+        },
+    }, {
+        'url': 'http://www.arte.tv/magazine/metropolis/de/frank-woeste-german-paris-metropolis',
+        'only_matching': True,
+    }]
+
+
+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)