Updated version 2017.09.24 from 'upstream/2017.09.24'

[youtubedl] / youtube_dl / extractor / youtube.py

diff --git a/youtube_dl/extractor/youtube.py b/youtube_dl/extractor/youtube.py
index 44a39282f104ff6bf588c7eee5122b190c8dba15..ad2e933ee4e34c9ebdb982ca66278e6e4c4a06b0 100644 (file)
--- a/youtube_dl/extractor/youtube.py
+++ b/youtube_dl/extractor/youtube.py
@@ -16,6 +16,7 @@ from ..jsinterp import JSInterpreter
 from ..swfinterp import SWFInterpreter
 from ..compat import (
     compat_chr,
 from ..swfinterp import SWFInterpreter
 from ..compat import (
     compat_chr,

+    compat_kwargs,

     compat_parse_qs,
     compat_urllib_parse_unquote,
     compat_urllib_parse_unquote_plus,
     compat_parse_qs,
     compat_urllib_parse_unquote,
     compat_urllib_parse_unquote_plus,


@@ -245,6 +246,11 @@ class YoutubeBaseInfoExtractor(InfoExtractor):
 
         return True
 
 
         return True
 

+    def _download_webpage(self, *args, **kwargs):
+        kwargs.setdefault('query', {})['disable_polymer'] = 'true'
+        return super(YoutubeBaseInfoExtractor, self)._download_webpage(
+            *args, **compat_kwargs(kwargs))
+

     def _real_initialize(self):
         if self._downloader is None:
             return
     def _real_initialize(self):
         if self._downloader is None:
             return


@@ -673,6 +679,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
             },
         },
         # video_info is None (https://github.com/rg3/youtube-dl/issues/4421)
             },
         },
         # video_info is None (https://github.com/rg3/youtube-dl/issues/4421)

+        # YouTube Red ad is not captured for creator

         {
             'url': '__2ABJjxzNo',
             'info_dict': {
         {
             'url': '__2ABJjxzNo',
             'info_dict': {


@@ -1002,6 +1009,27 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                 'Skipping DASH manifest',
             ],
         },
                 'Skipping DASH manifest',
             ],
         },

+        {
+            # The following content has been identified by the YouTube community
+            # as inappropriate or offensive to some audiences.
+            'url': 'https://www.youtube.com/watch?v=6SJNVb0GnPI',
+            'info_dict': {
+                'id': '6SJNVb0GnPI',
+                'ext': 'mp4',
+                'title': 'Race Differences in Intelligence',
+                'description': 'md5:5d161533167390427a1f8ee89a1fc6f1',
+                'duration': 965,
+                'upload_date': '20140124',
+                'uploader': 'New Century Foundation',
+                'uploader_id': 'UCEJYpZGqgUob0zVVEaLhvVg',
+                'uploader_url': r're:https?://(?:www\.)?youtube\.com/channel/UCEJYpZGqgUob0zVVEaLhvVg',
+                'license': 'Standard YouTube License',
+                'view_count': int,
+            },
+            'params': {
+                'skip_download': True,
+            },
+        },

         {
             # itag 212
             'url': '1t24XAntNCY',
         {
             # itag 212
             'url': '1t24XAntNCY',


@@ -1269,37 +1297,57 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                     sub_lang_list[sub_lang] = sub_formats
                 return sub_lang_list
 
                     sub_lang_list[sub_lang] = sub_formats
                 return sub_lang_list
 

+            def make_captions(sub_url, sub_langs):
+                parsed_sub_url = compat_urllib_parse_urlparse(sub_url)
+                caption_qs = compat_parse_qs(parsed_sub_url.query)
+                captions = {}
+                for sub_lang in sub_langs:
+                    sub_formats = []
+                    for ext in self._SUBTITLE_FORMATS:
+                        caption_qs.update({
+                            'tlang': [sub_lang],
+                            'fmt': [ext],
+                        })
+                        sub_url = compat_urlparse.urlunparse(parsed_sub_url._replace(
+                            query=compat_urllib_parse_urlencode(caption_qs, True)))
+                        sub_formats.append({
+                            'url': sub_url,
+                            'ext': ext,
+                        })
+                    captions[sub_lang] = sub_formats
+                return captions
+
+            # New captions format as of 22.06.2017
+            player_response = args.get('player_response')
+            if player_response and isinstance(player_response, compat_str):
+                player_response = self._parse_json(
+                    player_response, video_id, fatal=False)
+                if player_response:
+                    renderer = player_response['captions']['playerCaptionsTracklistRenderer']
+                    base_url = renderer['captionTracks'][0]['baseUrl']
+                    sub_lang_list = []
+                    for lang in renderer['translationLanguages']:
+                        lang_code = lang.get('languageCode')
+                        if lang_code:
+                            sub_lang_list.append(lang_code)
+                    return make_captions(base_url, sub_lang_list)
+

             # Some videos don't provide ttsurl but rather caption_tracks and
             # caption_translation_languages (e.g. 20LmZk1hakA)
             # Some videos don't provide ttsurl but rather caption_tracks and
             # caption_translation_languages (e.g. 20LmZk1hakA)

+            # Does not used anymore as of 22.06.2017

             caption_tracks = args['caption_tracks']
             caption_translation_languages = args['caption_translation_languages']
             caption_url = compat_parse_qs(caption_tracks.split(',')[0])['u'][0]
             caption_tracks = args['caption_tracks']
             caption_translation_languages = args['caption_translation_languages']
             caption_url = compat_parse_qs(caption_tracks.split(',')[0])['u'][0]

-            parsed_caption_url = compat_urllib_parse_urlparse(caption_url)
-            caption_qs = compat_parse_qs(parsed_caption_url.query)
-
-            sub_lang_list = {}
+            sub_lang_list = []

             for lang in caption_translation_languages.split(','):
                 lang_qs = compat_parse_qs(compat_urllib_parse_unquote_plus(lang))
                 sub_lang = lang_qs.get('lc', [None])[0]
             for lang in caption_translation_languages.split(','):
                 lang_qs = compat_parse_qs(compat_urllib_parse_unquote_plus(lang))
                 sub_lang = lang_qs.get('lc', [None])[0]

-                if not sub_lang:
-                    continue
-                sub_formats = []
-                for ext in self._SUBTITLE_FORMATS:
-                    caption_qs.update({
-                        'tlang': [sub_lang],
-                        'fmt': [ext],
-                    })
-                    sub_url = compat_urlparse.urlunparse(parsed_caption_url._replace(
-                        query=compat_urllib_parse_urlencode(caption_qs, True)))
-                    sub_formats.append({
-                        'url': sub_url,
-                        'ext': ext,
-                    })
-                sub_lang_list[sub_lang] = sub_formats
-            return sub_lang_list
+                if sub_lang:
+                    sub_lang_list.append(sub_lang)
+            return make_captions(caption_url, sub_lang_list)

         # An extractor error can be raise by the download process if there are
         # no automatic captions but there are subtitles
         # An extractor error can be raise by the download process if there are
         # no automatic captions but there are subtitles

-        except (KeyError, ExtractorError):
+        except (KeyError, IndexError, ExtractorError):

             self._downloader.report_warning(err_msg)
             return {}
 
             self._downloader.report_warning(err_msg)
             return {}
 


@@ -1326,6 +1374,43 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
             playback_url, video_id, 'Marking watched',
             'Unable to mark watched', fatal=False)
 
             playback_url, video_id, 'Marking watched',
             'Unable to mark watched', fatal=False)
 

+    @staticmethod
+    def _extract_urls(webpage):
+        # Embedded YouTube player
+        entries = [
+            unescapeHTML(mobj.group('url'))
+            for mobj in re.finditer(r'''(?x)
+            (?:
+                <iframe[^>]+?src=|
+                data-video-url=|
+                <embed[^>]+?src=|
+                embedSWF\(?:\s*|
+                <object[^>]+data=|
+                new\s+SWFObject\(
+            )
+            (["\'])
+                (?P<url>(?:https?:)?//(?:www\.)?youtube(?:-nocookie)?\.com/
+                (?:embed|v|p)/.+?)
+            \1''', webpage)]
+
+        # lazyYT YouTube embed
+        entries.extend(list(map(
+            unescapeHTML,
+            re.findall(r'class="lazyYT" data-youtube-id="([^"]+)"', webpage))))
+
+        # Wordpress "YouTube Video Importer" plugin
+        matches = re.findall(r'''(?x)<div[^>]+
+            class=(?P<q1>[\'"])[^\'"]*\byvii_single_video_player\b[^\'"]*(?P=q1)[^>]+
+            data-video_id=(?P<q2>[\'"])([^\'"]+)(?P=q2)''', webpage)
+        entries.extend(m[-1] for m in matches)
+
+        return entries
+
+    @staticmethod
+    def _extract_url(webpage):
+        urls = YoutubeIE._extract_urls(webpage)
+        return urls[0] if urls else None
+

     @classmethod
     def extract_id(cls, url):
         mobj = re.match(cls._VALID_URL, url, re.VERBOSE)
     @classmethod
     def extract_id(cls, url):
         mobj = re.match(cls._VALID_URL, url, re.VERBOSE)


@@ -1353,10 +1438,16 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
             start_time = parse_duration(time_point)
             if start_time is None:
                 continue
             start_time = parse_duration(time_point)
             if start_time is None:
                 continue

+            if start_time > duration:
+                break

             end_time = (duration if next_num == len(chapter_lines)
                         else parse_duration(chapter_lines[next_num][1]))
             if end_time is None:
                 continue
             end_time = (duration if next_num == len(chapter_lines)
                         else parse_duration(chapter_lines[next_num][1]))
             if end_time is None:
                 continue

+            if end_time > duration:
+                end_time = duration
+            if start_time > end_time:
+                break

             chapter_title = re.sub(
                 r'<a[^>]+>[^<]+</a>', '', chapter_line).strip(' \t-')
             chapter_title = re.sub(r'\s+', ' ', chapter_title)
             chapter_title = re.sub(
                 r'<a[^>]+>[^<]+</a>', '', chapter_line).strip(' \t-')
             chapter_title = re.sub(r'\s+', ' ', chapter_title)


@@ -1410,9 +1501,14 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
             if dash_mpd and dash_mpd[0] not in dash_mpds:
                 dash_mpds.append(dash_mpd[0])
 
             if dash_mpd and dash_mpd[0] not in dash_mpds:
                 dash_mpds.append(dash_mpd[0])
 

+        is_live = None
+        view_count = None
+
+        def extract_view_count(v_info):
+            return int_or_none(try_get(v_info, lambda x: x['view_count'][0]))
+

         # Get video info
         embed_webpage = None
         # Get video info
         embed_webpage = None

-        is_live = None

         if re.search(r'player-age-gate-content">', video_webpage) is not None:
             age_gate = True
             # We simulate the access to the video from www.youtube.com/v/{video_id}
         if re.search(r'player-age-gate-content">', video_webpage) is not None:
             age_gate = True
             # We simulate the access to the video from www.youtube.com/v/{video_id}


@@ -1435,6 +1531,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
         else:
             age_gate = False
             video_info = None
         else:
             age_gate = False
             video_info = None

+            sts = None

             # Try looking directly into the video webpage
             ytplayer_config = self._get_ytplayer_config(video_id, video_webpage)
             if ytplayer_config:
             # Try looking directly into the video webpage
             ytplayer_config = self._get_ytplayer_config(video_id, video_webpage)
             if ytplayer_config:


@@ -1451,6 +1548,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                         args['ypc_vid'], YoutubeIE.ie_key(), video_id=args['ypc_vid'])
                 if args.get('livestream') == '1' or args.get('live_playback') == 1:
                     is_live = True
                         args['ypc_vid'], YoutubeIE.ie_key(), video_id=args['ypc_vid'])
                 if args.get('livestream') == '1' or args.get('live_playback') == 1:
                     is_live = True

+                sts = ytplayer_config.get('sts')

             if not video_info or self._downloader.params.get('youtube_include_dash_manifest', True):
                 # We also try looking in get_video_info since it may contain different dashmpd
                 # URL that points to a DASH manifest with possibly different itag set (some itags
             if not video_info or self._downloader.params.get('youtube_include_dash_manifest', True):
                 # We also try looking in get_video_info since it may contain different dashmpd
                 # URL that points to a DASH manifest with possibly different itag set (some itags


@@ -1459,17 +1557,29 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                 # The general idea is to take a union of itags of both DASH manifests (for example
                 # video with such 'manifest behavior' see https://github.com/rg3/youtube-dl/issues/6093)
                 self.report_video_info_webpage_download(video_id)
                 # The general idea is to take a union of itags of both DASH manifests (for example
                 # video with such 'manifest behavior' see https://github.com/rg3/youtube-dl/issues/6093)
                 self.report_video_info_webpage_download(video_id)

-                for el_type in ['&el=info', '&el=embedded', '&el=detailpage', '&el=vevo', '']:
-                    video_info_url = (
-                        '%s://www.youtube.com/get_video_info?&video_id=%s%s&ps=default&eurl=&gl=US&hl=en'
-                        % (proto, video_id, el_type))
+                for el in ('info', 'embedded', 'detailpage', 'vevo', ''):
+                    query = {
+                        'video_id': video_id,
+                        'ps': 'default',
+                        'eurl': '',
+                        'gl': 'US',
+                        'hl': 'en',
+                    }
+                    if el:
+                        query['el'] = el
+                    if sts:
+                        query['sts'] = sts

                     video_info_webpage = self._download_webpage(
                     video_info_webpage = self._download_webpage(

-                        video_info_url,
+                        '%s://www.youtube.com/get_video_info' % proto,

                         video_id, note=False,
                         video_id, note=False,

-                        errnote='unable to download video info webpage')
+                        errnote='unable to download video info webpage',
+                        fatal=False, query=query)
+                    if not video_info_webpage:
+                        continue

                     get_video_info = compat_parse_qs(video_info_webpage)
                     get_video_info = compat_parse_qs(video_info_webpage)

-                    if get_video_info.get('use_cipher_signature') != ['True']:
-                        add_dash_mpd(get_video_info)
+                    add_dash_mpd(get_video_info)
+                    if view_count is None:
+                        view_count = extract_view_count(get_video_info)

                     if not video_info:
                         video_info = get_video_info
                     if 'token' in get_video_info:
                     if not video_info:
                         video_info = get_video_info
                     if 'token' in get_video_info:


@@ -1553,10 +1663,8 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                 return self.playlist_result(entries, video_id, video_title, video_description)
             self.to_screen('Downloading just video %s because of --no-playlist' % video_id)
 
                 return self.playlist_result(entries, video_id, video_title, video_description)
             self.to_screen('Downloading just video %s because of --no-playlist' % video_id)
 

-        if 'view_count' in video_info:
-            view_count = int(video_info['view_count'][0])
-        else:
-            view_count = None
+        if view_count is None:
+            view_count = extract_view_count(video_info)

 
         # Check for "rental" videos
         if 'ypc_video_rental_bar_text' in video_info and 'author' not in video_info:
 
         # Check for "rental" videos
         if 'ypc_video_rental_bar_text' in video_info and 'author' not in video_info:


@@ -1600,10 +1708,8 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
         if not upload_date:
             upload_date = self._search_regex(
                 [r'(?s)id="eow-date.*?>(.*?)</span>',
         if not upload_date:
             upload_date = self._search_regex(
                 [r'(?s)id="eow-date.*?>(.*?)</span>',

-                 r'id="watch-uploader-info".*?>.*?(?:Published|Uploaded|Streamed live|Started) on (.+?)</strong>'],
+                 r'(?:id="watch-uploader-info".*?>.*?|["\']simpleText["\']\s*:\s*["\'])(?:Published|Uploaded|Streamed live|Started) on (.+?)[<"\']'],

                 video_webpage, 'upload date', default=None)
                 video_webpage, 'upload date', default=None)

-            if upload_date:
-                upload_date = ' '.join(re.sub(r'[/,-]', r' ', mobj.group(1)).split())

         upload_date = unified_strdate(upload_date)
 
         video_license = self._html_search_regex(
         upload_date = unified_strdate(upload_date)
 
         video_license = self._html_search_regex(


@@ -1611,7 +1717,21 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
             video_webpage, 'license', default=None)
 
         m_music = re.search(
             video_webpage, 'license', default=None)
 
         m_music = re.search(

-            r'<h4[^>]+class="title"[^>]*>\s*Music\s*</h4>\s*<ul[^>]*>\s*<li>(?P<title>.+?) by (?P<creator>.+?)(?:\(.+?\))?</li',
+            r'''(?x)
+                <h4[^>]+class="title"[^>]*>\s*Music\s*</h4>\s*
+                <ul[^>]*>\s*
+                <li>(?P<title>.+?)
+                by (?P<creator>.+?)
+                (?:
+                    \(.+?\)|
+                    <a[^>]*
+                        (?:
+                            \bhref=["\']/red[^>]*>|             # drop possible
+                            >\s*Listen ad-free with YouTube Red # YouTube Red ad
+                        )
+                    .*?
+                )?</li
+            ''',

             video_webpage)
         if m_music:
             video_alt_title = remove_quotes(unescapeHTML(m_music.group('title')))
             video_webpage)
         if m_music:
             video_alt_title = remove_quotes(unescapeHTML(m_music.group('title')))


@@ -1703,12 +1823,8 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                 format_id = url_data['itag'][0]
                 url = url_data['url'][0]
 
                 format_id = url_data['itag'][0]
                 url = url_data['url'][0]
 

-                if 'sig' in url_data:
-                    url += '&signature=' + url_data['sig'][0]
-                elif 's' in url_data:
-                    encrypted_sig = url_data['s'][0]
+                if 's' in url_data or self._downloader.params.get('youtube_include_dash_manifest', True):

                     ASSETS_RE = r'"assets":.+?"js":\s*("[^"]+")'
                     ASSETS_RE = r'"assets":.+?"js":\s*("[^"]+")'

-

                     jsplayer_url_json = self._search_regex(
                         ASSETS_RE,
                         embed_webpage if age_gate else video_webpage,
                     jsplayer_url_json = self._search_regex(
                         ASSETS_RE,
                         embed_webpage if age_gate else video_webpage,


@@ -1729,6 +1845,11 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                             video_webpage, 'age gate player URL')
                         player_url = json.loads(player_url_json)
 
                             video_webpage, 'age gate player URL')
                         player_url = json.loads(player_url_json)
 

+                if 'sig' in url_data:
+                    url += '&signature=' + url_data['sig'][0]
+                elif 's' in url_data:
+                    encrypted_sig = url_data['s'][0]
+

                     if self._downloader.params.get('verbose'):
                         if player_url is None:
                             player_version = 'unknown'
                     if self._downloader.params.get('verbose'):
                         if player_url is None:
                             player_version = 'unknown'


@@ -1974,7 +2095,7 @@ class YoutubePlaylistIE(YoutubePlaylistBaseInfoExtractor):
                      |
                         (%(playlist_id)s)
                      )""" % {'playlist_id': YoutubeBaseInfoExtractor._PLAYLIST_ID_RE}
                      |
                         (%(playlist_id)s)
                      )""" % {'playlist_id': YoutubeBaseInfoExtractor._PLAYLIST_ID_RE}

-    _TEMPLATE_URL = 'https://www.youtube.com/playlist?list=%s&disable_polymer=true'
+    _TEMPLATE_URL = 'https://www.youtube.com/playlist?list=%s'

     _VIDEO_RE = r'href="\s*/watch\?v=(?P<id>[0-9A-Za-z_-]{11})&amp;[^"]*?index=(?P<index>\d+)(?:[^>]+>(?P<title>[^<]+))?'
     IE_NAME = 'youtube:playlist'
     _TESTS = [{
     _VIDEO_RE = r'href="\s*/watch\?v=(?P<id>[0-9A-Za-z_-]{11})&amp;[^"]*?index=(?P<index>\d+)(?:[^>]+>(?P<title>[^<]+))?'
     IE_NAME = 'youtube:playlist'
     _TESTS = [{