]> Raphaƫl G. Git Repositories - youtubedl/blobdiff - youtube_dl/extractor/youtube.py
Prepate to upload
[youtubedl] / youtube_dl / extractor / youtube.py
index 9943dddc13b478ce23e71c43d19f5cf18694f90b..e7bd1f18fb504c8541d0fbdc301980cab07d1f54 100644 (file)
@@ -87,7 +87,7 @@ class YoutubeBaseInfoExtractor(InfoExtractor):
         (username, password) = self._get_login_info()
         # No authentication to be performed
         if username is None:
-            if self._LOGIN_REQUIRED:
+            if self._LOGIN_REQUIRED and self._downloader.params.get('cookiefile') is None:
                 raise ExtractorError('No login info available, needed for using %s.' % self.IE_NAME, expected=True)
             return True
 
@@ -1596,6 +1596,12 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                         if 'token' not in video_info:
                             video_info = get_video_info
                         break
+
+        def extract_unavailable_message():
+            return self._html_search_regex(
+                r'(?s)<h1[^>]+id="unavailable-message"[^>]*>(.+?)</h1>',
+                video_webpage, 'unavailable message', default=None)
+
         if 'token' not in video_info:
             if 'reason' in video_info:
                 if 'The uploader has not made this video available in your country.' in video_info['reason']:
@@ -1604,8 +1610,13 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                     countries = regions_allowed.split(',') if regions_allowed else None
                     self.raise_geo_restricted(
                         msg=video_info['reason'][0], countries=countries)
+                reason = video_info['reason'][0]
+                if 'Invalid parameters' in reason:
+                    unavailable_message = extract_unavailable_message()
+                    if unavailable_message:
+                        reason = unavailable_message
                 raise ExtractorError(
-                    'YouTube said: %s' % video_info['reason'][0],
+                    'YouTube said: %s' % reason,
                     expected=True, video_id=video_id)
             else:
                 raise ExtractorError(
@@ -1810,7 +1821,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                 'url': video_info['conn'][0],
                 'player_url': player_url,
             }]
-        elif len(video_info.get('url_encoded_fmt_stream_map', [''])[0]) >= 1 or len(video_info.get('adaptive_fmts', [''])[0]) >= 1:
+        elif not is_live and (len(video_info.get('url_encoded_fmt_stream_map', [''])[0]) >= 1 or len(video_info.get('adaptive_fmts', [''])[0]) >= 1):
             encoded_url_map = video_info.get('url_encoded_fmt_stream_map', [''])[0] + ',' + video_info.get('adaptive_fmts', [''])[0]
             if 'rtmpe%3Dyes' in encoded_url_map:
                 raise ExtractorError('rtmpe downloads are not supported, see https://github.com/rg3/youtube-dl/issues/343 for more information.', expected=True)
@@ -1933,6 +1944,11 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                                     break
                             if codecs:
                                 dct.update(parse_codecs(codecs))
+                if dct.get('acodec') == 'none' or dct.get('vcodec') == 'none':
+                    dct['downloader_options'] = {
+                        # Youtube throttles chunks >~10M
+                        'http_chunk_size': 10485760,
+                    }
                 formats.append(dct)
         elif video_info.get('hlsvp'):
             manifest_url = video_info['hlsvp'][0]
@@ -1953,9 +1969,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                 a_format.setdefault('http_headers', {})['Youtubedl-no-compression'] = 'True'
                 formats.append(a_format)
         else:
-            unavailable_message = self._html_search_regex(
-                r'(?s)<h1[^>]+id="unavailable-message"[^>]*>(.+?)</h1>',
-                video_webpage, 'unavailable message', default=None)
+            unavailable_message = extract_unavailable_message()
             if unavailable_message:
                 raise ExtractorError(unavailable_message, expected=True)
             raise ExtractorError('no conn, hlsvp or url_encoded_fmt_stream_map information found in video info')
@@ -2270,6 +2284,19 @@ class YoutubePlaylistIE(YoutubePlaylistBaseInfoExtractor):
             r'(?s)<h1 class="pl-header-title[^"]*"[^>]*>\s*(.*?)\s*</h1>',
             page, 'title', default=None)
 
+        _UPLOADER_BASE = r'class=["\']pl-header-details[^>]+>\s*<li>\s*<a[^>]+\bhref='
+        uploader = self._search_regex(
+            r'%s["\']/(?:user|channel)/[^>]+>([^<]+)' % _UPLOADER_BASE,
+            page, 'uploader', default=None)
+        mobj = re.search(
+            r'%s(["\'])(?P<path>/(?:user|channel)/(?P<uploader_id>.+?))\1' % _UPLOADER_BASE,
+            page)
+        if mobj:
+            uploader_id = mobj.group('uploader_id')
+            uploader_url = compat_urlparse.urljoin(url, mobj.group('path'))
+        else:
+            uploader_id = uploader_url = None
+
         has_videos = True
 
         if not playlist_title:
@@ -2280,8 +2307,15 @@ class YoutubePlaylistIE(YoutubePlaylistBaseInfoExtractor):
             except StopIteration:
                 has_videos = False
 
-        return has_videos, self.playlist_result(
+        playlist = self.playlist_result(
             self._entries(page, playlist_id), playlist_id, playlist_title)
+        playlist.update({
+            'uploader': uploader,
+            'uploader_id': uploader_id,
+            'uploader_url': uploader_url,
+        })
+
+        return has_videos, playlist
 
     def _check_download_just_video(self, url, playlist_id):
         # Check if it's a video-specific URL
@@ -2417,7 +2451,7 @@ class YoutubeChannelIE(YoutubePlaylistBaseInfoExtractor):
 
 class YoutubeUserIE(YoutubeChannelIE):
     IE_DESC = 'YouTube.com user videos (URL or "ytuser" keyword)'
-    _VALID_URL = r'(?:(?:https?://(?:\w+\.)?youtube\.com/(?:(?P<user>user|c)/)?(?!(?:attribution_link|watch|results)(?:$|[^a-z_A-Z0-9-])))|ytuser:)(?!feed/)(?P<id>[A-Za-z0-9_-]+)'
+    _VALID_URL = r'(?:(?:https?://(?:\w+\.)?youtube\.com/(?:(?P<user>user|c)/)?(?!(?:attribution_link|watch|results|shared)(?:$|[^a-z_A-Z0-9-])))|ytuser:)(?!feed/)(?P<id>[A-Za-z0-9_-]+)'
     _TEMPLATE_URL = 'https://www.youtube.com/%s/%s/videos'
     IE_NAME = 'youtube:user'
 
@@ -2510,10 +2544,11 @@ class YoutubeLiveIE(YoutubeBaseInfoExtractor):
         webpage = self._download_webpage(url, channel_id, fatal=False)
         if webpage:
             page_type = self._og_search_property(
-                'type', webpage, 'page type', default=None)
+                'type', webpage, 'page type', default='')
             video_id = self._html_search_meta(
                 'videoId', webpage, 'video id', default=None)
-            if page_type == 'video' and video_id and re.match(r'^[0-9A-Za-z_-]{11}$', video_id):
+            if page_type.startswith('video') and video_id and re.match(
+                    r'^[0-9A-Za-z_-]{11}$', video_id):
                 return self.url_result(video_id, YoutubeIE.ie_key())
         return self.url_result(base_url)
 
@@ -2548,7 +2583,11 @@ class YoutubePlaylistsIE(YoutubePlaylistsBaseInfoExtractor):
     }]
 
 
-class YoutubeSearchIE(SearchInfoExtractor, YoutubePlaylistIE):
+class YoutubeSearchBaseInfoExtractor(YoutubePlaylistBaseInfoExtractor):
+    _VIDEO_RE = r'href="\s*/watch\?v=(?P<id>[0-9A-Za-z_-]{11})(?:[^"]*"[^>]+\btitle="(?P<title>[^"]+))?'
+
+
+class YoutubeSearchIE(SearchInfoExtractor, YoutubeSearchBaseInfoExtractor):
     IE_DESC = 'YouTube.com searches'
     # there doesn't appear to be a real limit, for example if you search for
     # 'python' you get more than 8.000.000 results
@@ -2582,8 +2621,7 @@ class YoutubeSearchIE(SearchInfoExtractor, YoutubePlaylistIE):
                 raise ExtractorError(
                     '[youtube] No video results', expected=True)
 
-            new_videos = self._ids_to_results(orderedSet(re.findall(
-                r'href="/watch\?v=(.{11})', html_content)))
+            new_videos = list(self._process_page(html_content))
             videos += new_videos
             if not new_videos or len(videos) > limit:
                 break
@@ -2606,11 +2644,10 @@ class YoutubeSearchDateIE(YoutubeSearchIE):
     _EXTRA_QUERY_ARGS = {'search_sort': 'video_date_uploaded'}
 
 
-class YoutubeSearchURLIE(YoutubePlaylistBaseInfoExtractor):
+class YoutubeSearchURLIE(YoutubeSearchBaseInfoExtractor):
     IE_DESC = 'YouTube.com search URLs'
     IE_NAME = 'youtube:search_url'
     _VALID_URL = r'https?://(?:www\.)?youtube\.com/results\?(.*?&)?(?:search_query|q)=(?P<query>[^&]+)(?:[&]|$)'
-    _VIDEO_RE = r'href="\s*/watch\?v=(?P<id>[0-9A-Za-z_-]{11})(?:[^"]*"[^>]+\btitle="(?P<title>[^"]+))?'
     _TESTS = [{
         'url': 'https://www.youtube.com/results?baz=bar&search_query=youtube-dl+test+video&filters=video&lclk=video',
         'playlist_mincount': 5,
@@ -2662,10 +2699,7 @@ class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor):
     def _real_initialize(self):
         self._login()
 
-    def _real_extract(self, url):
-        page = self._download_webpage(
-            'https://www.youtube.com/feed/%s' % self._FEED_NAME, self._PLAYLIST_TITLE)
-
+    def _entries(self, page):
         # The extraction process is the same as for playlists, but the regex
         # for the video ids doesn't contain an index
         ids = []
@@ -2676,12 +2710,15 @@ class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor):
             # 'recommended' feed has infinite 'load more' and each new portion spins
             # the same videos in (sometimes) slightly different order, so we'll check
             # for unicity and break when portion has no new videos
-            new_ids = filter(lambda video_id: video_id not in ids, orderedSet(matches))
+            new_ids = list(filter(lambda video_id: video_id not in ids, orderedSet(matches)))
             if not new_ids:
                 break
 
             ids.extend(new_ids)
 
+            for entry in self._ids_to_results(new_ids):
+                yield entry
+
             mobj = re.search(r'data-uix-load-more-href="/?(?P<more>[^"]+)"', more_widget_html)
             if not mobj:
                 break
@@ -2693,8 +2730,12 @@ class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor):
             content_html = more['content_html']
             more_widget_html = more['load_more_widget_html']
 
+    def _real_extract(self, url):
+        page = self._download_webpage(
+            'https://www.youtube.com/feed/%s' % self._FEED_NAME,
+            self._PLAYLIST_TITLE)
         return self.playlist_result(
-            self._ids_to_results(ids), playlist_title=self._PLAYLIST_TITLE)
+            self._entries(page), playlist_title=self._PLAYLIST_TITLE)
 
 
 class YoutubeWatchLaterIE(YoutubePlaylistIE):