]> Raphaƫl G. Git Repositories - youtubedl/blobdiff - youtube_dl/extractor/arte.py
Prepare to upload
[youtubedl] / youtube_dl / extractor / arte.py
index 69b3b0ad7820600ef5107ad3d79230c0e4edcaac..2bd3bfe8a852159cf7aefa93e731bab995ca7d23 100644 (file)
+# coding: utf-8
+from __future__ import unicode_literals
+
 import re
-import json
-import xml.etree.ElementTree
 
 from .common import InfoExtractor
+from ..compat import compat_str
 from ..utils import (
     ExtractorError,
-    find_xpath_attr,
+    int_or_none,
+    qualities,
+    try_get,
     unified_strdate,
 )
 
-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'
-    _LIVEWEB_URL = r'(?:http://)?liveweb.arte.tv/(?P<lang>fr|de)/(?P<subpage>.+?)/(?P<name>.+)'
-    _LIVE_URL = r'index-[0-9]+\.html$'
-
-    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, cls._LIVEWEB_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'))
+# 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.
 
-    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)
-
-        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, video_id) 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)
+
+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']
 
-        info_dict = {'id': player_info['VID'],
-                     'title': player_info['VTI'],
-                     'description': player_info.get('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']))
-        # Prefer videos without subtitles in the same language
-        formats = sorted(formats, key=lambda f: re.match(r'VO(F|A)-STM\1', f['versionCode']) is None)
-        # 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']
+        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.get('VTI') or title 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]',
+        }
+
+        langcode = LANGS.get(lang, lang)
+
+        formats = []
+        for format_id, format_dict in vsr.items():
+            f = dict(format_dict)
+            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
+
+            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')),
+            }
+
+            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
 
-    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')
-
-        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
-            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',
-                }
-
-    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, u'event id')
-        config_xml = self._download_webpage('http://download.liveweb.arte.tv/o21/liveweb/events/event-%s.xml' % video_id,
-                                            video_id, u'Downloading information')
-        config_doc = xml.etree.ElementTree.fromstring(config_xml.encode('utf-8'))
-        event_doc = config_doc.find('event')
-        url_node = event_doc.find('video').find('urlHd')
-        if url_node is None:
-            url_node = video_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(ArteTVBaseIE):
+    IE_NAME = 'arte.tv:+7'
+    _VALID_URL = r'https?://(?:www\.)?arte\.tv/(?P<lang>fr|de|en|es|it|pl)/videos/(?P<id>\d{6}-\d{3}-[AF])'
+
+    _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',
+        },
+    }]
+
+    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 ArteTVEmbedIE(ArteTVPlus7IE):
+    IE_NAME = 'arte.tv:embed'
+    _VALID_URL = r'''(?x)
+        https://www\.arte\.tv
+        /player/v3/index\.php\?json_url=
+        (?P<json_url>
+            https?://api\.arte\.tv/api/player/v1/config/
+            (?P<lang>[^/]+)/(?P<id>\d{6}-\d{3}-[AF])
+        )
+    '''
+
+    _TESTS = []
+
+    def _real_extract(self, 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 ArteTVPlaylistIE(ArteTVBaseIE):
+    IE_NAME = 'arte.tv:playlist'
+    _VALID_URL = r'https?://(?:www\.)?arte\.tv/(?P<lang>fr|de|en|es|it|pl)/videos/(?P<id>RC-\d{6})'
+
+    _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,
+    }]
+
+    def _real_extract(self, url):
+        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)