from ..swfinterp import SWFInterpreter
from ..compat import (
compat_chr,
+ compat_kwargs,
compat_parse_qs,
compat_urllib_parse_unquote,
compat_urllib_parse_unquote_plus,
(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
return True
+ def _download_webpage_handle(self, *args, **kwargs):
+ kwargs.setdefault('query', {})['disable_polymer'] = 'true'
+ return super(YoutubeBaseInfoExtractor, self)._download_webpage_handle(
+ *args, **compat_kwargs(kwargs))
+
def _real_initialize(self):
if self._downloader is None:
return
(?:(?:(?:(?:\w+\.)?[yY][oO][uU][tT][uU][bB][eE](?:-nocookie)?\.com/|
(?:www\.)?deturl\.com/www\.youtube\.com/|
(?:www\.)?pwnyoutube\.com/|
+ (?:www\.)?hooktube\.com/|
(?:www\.)?yourepeat\.com/|
tube\.majestyc\.net/|
youtube\.googleapis\.com/) # the various hostnames, with wildcard subdomains
},
},
# video_info is None (https://github.com/rg3/youtube-dl/issues/4421)
+ # YouTube Red ad is not captured for creator
{
'url': '__2ABJjxzNo',
'info_dict': {
'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',
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)
+ # 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]
- 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]
- 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
- except (KeyError, ExtractorError):
+ except (KeyError, IndexError, ExtractorError):
self._downloader.report_warning(err_msg)
return {}
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)/[0-9A-Za-z_-]{11}.*?)
+ \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)
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
- 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}
ytplayer_config = self._get_ytplayer_config(video_id, video_webpage)
if ytplayer_config:
args = ytplayer_config['args']
- if args.get('url_encoded_fmt_stream_map'):
+ if args.get('url_encoded_fmt_stream_map') or args.get('hlsvp'):
# Convert to the same format returned by compat_parse_qs
video_info = dict((k, [v]) for k, v in args.items())
add_dash_mpd(video_info)
continue
get_video_info = compat_parse_qs(video_info_webpage)
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 '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']:
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(
# description
description_original = video_description = get_element_by_id("eow-description", video_webpage)
if video_description:
+
+ def replace_url(m):
+ redir_url = compat_urlparse.urljoin(url, m.group(1))
+ parsed_redir_url = compat_urllib_parse_urlparse(redir_url)
+ if re.search(r'^(?:www\.)?(?:youtube(?:-nocookie)?\.com|youtu\.be)$', parsed_redir_url.netloc) and parsed_redir_url.path == '/redirect':
+ qs = compat_parse_qs(parsed_redir_url.query)
+ q = qs.get('q')
+ if q and q[0]:
+ return q[0]
+ return redir_url
+
description_original = video_description = re.sub(r'''(?x)
<a\s+
(?:[a-zA-Z-]+="[^"]*"\s+)*?
class="[^"]*"[^>]*>
[^<]+\.{3}\s*
</a>
- ''', r'\1', video_description)
+ ''', replace_url, video_description)
video_description = clean_html(video_description)
else:
fd_mobj = re.search(r'<meta name="description" content="([^"]+)"', video_webpage)
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:
self.report_information_extraction(video_id)
# uploader
- if 'author' not in video_info:
- raise ExtractorError('Unable to extract uploader name')
- video_uploader = compat_urllib_parse_unquote_plus(video_info['author'][0])
+ video_uploader = try_get(video_info, lambda x: x['author'][0], compat_str)
+ if video_uploader:
+ video_uploader = compat_urllib_parse_unquote_plus(video_uploader)
+ else:
+ self._downloader.report_warning('unable to extract uploader name')
# uploader_id
video_uploader_id = None
video_uploader_url = None
mobj = re.search(
- r'<link itemprop="url" href="(?P<uploader_url>https?://www.youtube.com/(?:user|channel)/(?P<uploader_id>[^"]+))">',
+ r'<link itemprop="url" href="(?P<uploader_url>https?://www\.youtube\.com/(?:user|channel)/(?P<uploader_id>[^"]+))">',
video_webpage)
if mobj is not None:
video_uploader_id = mobj.group('uploader_id')
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)
- 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(
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')))
chapters = self._extract_chapters(description_original, video_duration)
+ def _extract_filesize(media_url):
+ return int_or_none(self._search_regex(
+ r'\bclen[=/](\d+)', media_url, 'filesize', default=None))
+
if 'conn' in video_info and video_info['conn'][0].startswith('rtmp'):
self.report_rtmp_download()
formats = [{
'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)
mobj = re.search(r'^(?P<width>\d+)[xX](?P<height>\d+)$', url_data.get('size', [''])[0])
width, height = (int(mobj.group('width')), int(mobj.group('height'))) if mobj else (None, None)
+ filesize = int_or_none(url_data.get(
+ 'clen', [None])[0]) or _extract_filesize(url)
+
more_fields = {
- 'filesize': int_or_none(url_data.get('clen', [None])[0]),
+ 'filesize': filesize,
'tbr': float_or_none(url_data.get('bitrate', [None])[0], 1000),
'width': width,
'height': height,
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]
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)
- if unavailable_message:
- raise ExtractorError(unavailable_message, expected=True)
+ error_message = clean_html(video_info.get('reason', [None])[0])
+ if not error_message:
+ error_message = extract_unavailable_message()
+ if error_message:
+ raise ExtractorError(error_message, expected=True)
raise ExtractorError('no conn, hlsvp or url_encoded_fmt_stream_map information found in video info')
# Look for the DASH manifest
for df in self._extract_mpd_formats(
mpd_url, video_id, fatal=dash_mpd_fatal,
formats_dict=self._formats):
+ if not df.get('filesize'):
+ df['filesize'] = _extract_filesize(df['url'])
# Do not overwrite DASH format found in some previous DASH manifest
if df['format_id'] not in dash_formats:
dash_formats[df['format_id']] = df
}
-class YoutubeSharedVideoIE(InfoExtractor):
- _VALID_URL = r'(?:https?:)?//(?:www\.)?youtube\.com/shared\?.*\bci=(?P<id>[0-9A-Za-z_-]{11})'
- IE_NAME = 'youtube:shared'
-
- _TEST = {
- 'url': 'https://www.youtube.com/shared?ci=1nEzmT-M4fU',
- 'info_dict': {
- 'id': 'uPDB5I9wfp8',
- 'ext': 'webm',
- 'title': 'Pocoyo: 90 minutos de episódios completos Português para crianças - PARTE 3',
- 'description': 'md5:d9e4d9346a2dfff4c7dc4c8cec0f546d',
- 'upload_date': '20160219',
- 'uploader': 'Pocoyo - Português (BR)',
- 'uploader_id': 'PocoyoBrazil',
- },
- 'add_ie': ['Youtube'],
- 'params': {
- # There are already too many Youtube downloads
- 'skip_download': True,
- },
- }
-
- def _real_extract(self, url):
- video_id = self._match_id(url)
-
- webpage = self._download_webpage(url, video_id)
-
- real_video_id = self._html_search_meta(
- 'videoId', webpage, 'YouTube video id', fatal=True)
-
- return self.url_result(real_video_id, YoutubeIE.ie_key())
-
-
class YoutubePlaylistIE(YoutubePlaylistBaseInfoExtractor):
IE_DESC = 'YouTube.com playlists'
_VALID_URL = r"""(?x)(?:
|
(%(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})&[^"]*?index=(?P<index>\d+)(?:[^>]+>(?P<title>[^<]+))?'
IE_NAME = 'youtube:playlist'
_TESTS = [{
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:
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
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)(?:$|[^a-z_A-Z0-9-])))|ytuser:)(?!feed/)(?P<id>[A-Za-z0-9_-]+)'
_TEMPLATE_URL = 'https://www.youtube.com/%s/%s/videos'
IE_NAME = 'youtube:user'
webpage = self._download_webpage(url, channel_id, fatal=False)
if webpage:
page_type = self._og_search_property(
- 'type', webpage, 'page type', default=None)
+ 'type', webpage, 'page type', default='')
video_id = self._html_search_meta(
'videoId', webpage, 'video id', default=None)
- if page_type == 'video' and video_id and re.match(r'^[0-9A-Za-z_-]{11}$', video_id):
+ if page_type.startswith('video') and video_id and re.match(
+ r'^[0-9A-Za-z_-]{11}$', video_id):
return self.url_result(video_id, YoutubeIE.ie_key())
return self.url_result(base_url)
}]
-class YoutubeSearchIE(SearchInfoExtractor, YoutubePlaylistIE):
+class YoutubeSearchBaseInfoExtractor(YoutubePlaylistBaseInfoExtractor):
+ _VIDEO_RE = r'href="\s*/watch\?v=(?P<id>[0-9A-Za-z_-]{11})(?:[^"]*"[^>]+\btitle="(?P<title>[^"]+))?'
+
+
+class YoutubeSearchIE(SearchInfoExtractor, YoutubeSearchBaseInfoExtractor):
IE_DESC = 'YouTube.com searches'
# there doesn't appear to be a real limit, for example if you search for
# 'python' you get more than 8.000.000 results
raise ExtractorError(
'[youtube] No video results', expected=True)
- new_videos = self._ids_to_results(orderedSet(re.findall(
- r'href="/watch\?v=(.{11})', html_content)))
+ new_videos = list(self._process_page(html_content))
videos += new_videos
if not new_videos or len(videos) > limit:
break
_EXTRA_QUERY_ARGS = {'search_sort': 'video_date_uploaded'}
-class YoutubeSearchURLIE(YoutubePlaylistBaseInfoExtractor):
+class YoutubeSearchURLIE(YoutubeSearchBaseInfoExtractor):
IE_DESC = 'YouTube.com search URLs'
IE_NAME = 'youtube:search_url'
_VALID_URL = r'https?://(?:www\.)?youtube\.com/results\?(.*?&)?(?:search_query|q)=(?P<query>[^&]+)(?:[&]|$)'
- _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,
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 = []
# '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
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):
projects / youtube-dl / blobdiff