import string
import struct
import traceback
-import xml.etree.ElementTree
import zlib
from .common import InfoExtractor, SearchInfoExtractor
clean_html,
get_cachedir,
get_element_by_id,
+ get_element_by_attribute,
ExtractorError,
unescapeHTML,
unified_strdate,
'248': 'webm',
}
_video_dimensions = {
- '5': '240x400',
+ '5': '400x240',
'6': '???',
'13': '???',
- '17': '144x176',
- '18': '360x640',
- '22': '720x1280',
- '34': '360x640',
- '35': '480x854',
- '36': '240x320',
- '37': '1080x1920',
- '38': '3072x4096',
- '43': '360x640',
- '44': '480x854',
- '45': '720x1280',
- '46': '1080x1920',
+ '17': '176x144',
+ '18': '640x360',
+ '22': '1280x720',
+ '34': '640x360',
+ '35': '854x480',
+ '36': '320x240',
+ '37': '1920x1080',
+ '38': '4096x3072',
+ '43': '640x360',
+ '44': '854x480',
+ '45': '1280x720',
+ '46': '1920x1080',
'82': '360p',
'83': '480p',
'84': '720p',
u"uploader": u"Philipp Hagemeister",
u"uploader_id": u"phihag",
u"upload_date": u"20121002",
- u"description": u"test chars: \"'/\\ä↭𝕐\n\nThis is a test video for youtube-dl.\n\nFor more information, contact phihag@phihag.de ."
+ u"description": u"test chars: \"'/\\ä↭𝕐\ntest URL: https://github.com/rg3/youtube-dl/issues/1892\n\nThis is a test video for youtube-dl.\n\nFor more information, contact phihag@phihag.de ."
}
},
{
'asrs': 1,
})
list_url = caption_url + '&' + list_params
- list_page = self._download_webpage(list_url, video_id)
- caption_list = xml.etree.ElementTree.fromstring(list_page.encode('utf-8'))
+ caption_list = self._download_xml(list_url, video_id)
original_lang_node = caption_list.find('track')
if original_lang_node is None or original_lang_node.attrib.get('kind') != 'asr' :
self._downloader.report_warning(u'Video doesn\'t have automatic captions')
# description
video_description = get_element_by_id("eow-description", video_webpage)
if video_description:
+ video_description = re.sub(r'''(?x)
+ <a\s+
+ (?:[a-zA-Z-]+="[^"]+"\s+)*?
+ title="([^"]+)"\s+
+ (?:[a-zA-Z-]+="[^"]+"\s+)*?
+ class="yt-uix-redirect-link"\s*>
+ [^<]+
+ </a>
+ ''', r'\1', video_description)
video_description = clean_html(video_description)
else:
fd_mobj = re.search(r'<meta name="description" content="([^"]+)"', video_webpage)
})
return results
-class YoutubePlaylistIE(InfoExtractor):
+class YoutubePlaylistIE(YoutubeBaseInfoExtractor):
IE_DESC = u'YouTube.com playlists'
_VALID_URL = r"""(?:
(?:https?://)?
|
((?:PL|EC|UU|FL)[0-9A-Za-z-_]{10,})
)"""
- _TEMPLATE_URL = 'https://gdata.youtube.com/feeds/api/playlists/%s?max-results=%i&start-index=%i&v=2&alt=json&safeSearch=none'
- _MAX_RESULTS = 50
+ _TEMPLATE_URL = 'https://www.youtube.com/playlist?list=%s&page=%s'
+ _MORE_PAGES_INDICATOR = r'data-link-type="next"'
+ _VIDEO_RE = r'href="/watch\?v=(?P<id>[0-9A-Za-z_-]{11})&[^"]*?index=(?P<index>\d+)'
IE_NAME = u'youtube:playlist'
@classmethod
"""Receives a URL and returns True if suitable for this IE."""
return re.match(cls._VALID_URL, url, re.VERBOSE) is not None
+ def _real_initialize(self):
+ self._login()
+
+ def _ids_to_results(self, ids):
+ return [self.url_result(vid_id, 'Youtube', video_id=vid_id)
+ for vid_id in ids]
+
+ def _extract_mix(self, playlist_id):
+ # The mixes are generated from a a single video
+ # the id of the playlist is just 'RD' + video_id
+ url = 'https://youtube.com/watch?v=%s&list=%s' % (playlist_id[2:], playlist_id)
+ webpage = self._download_webpage(url, playlist_id, u'Downloading Youtube mix')
+ title_span = (get_element_by_attribute('class', 'title long-title', webpage) or
+ get_element_by_attribute('class', 'title ', webpage))
+ title = clean_html(title_span)
+ video_re = r'data-index="\d+".*?href="/watch\?v=([0-9A-Za-z_-]{11})&[^"]*?list=%s' % re.escape(playlist_id)
+ ids = orderedSet(re.findall(video_re, webpage))
+ url_results = self._ids_to_results(ids)
+
+ return self.playlist_result(url_results, playlist_id, title)
+
def _real_extract(self, url):
# Extract playlist id
mobj = re.match(self._VALID_URL, url, re.VERBOSE)
video_id = query_dict['v'][0]
if self._downloader.params.get('noplaylist'):
self.to_screen(u'Downloading just video %s because of --no-playlist' % video_id)
- return self.url_result('https://www.youtube.com/watch?v=' + video_id, 'Youtube')
+ return self.url_result(video_id, 'Youtube', video_id=video_id)
else:
self.to_screen(u'Downloading playlist PL%s - add --no-playlist to just download video %s' % (playlist_id, video_id))
- # Download playlist videos from API
- videos = []
+ if len(playlist_id) == 13: # 'RD' + 11 characters for the video id
+ # Mixes require a custom extraction process
+ return self._extract_mix(playlist_id)
+
+ # Extract the video ids from the playlist pages
+ ids = []
for page_num in itertools.count(1):
- start_index = self._MAX_RESULTS * (page_num - 1) + 1
- if start_index >= 1000:
- self._downloader.report_warning(u'Max number of results reached')
- break
- url = self._TEMPLATE_URL % (playlist_id, self._MAX_RESULTS, start_index)
+ url = self._TEMPLATE_URL % (playlist_id, page_num)
page = self._download_webpage(url, playlist_id, u'Downloading page #%s' % page_num)
+ matches = re.finditer(self._VIDEO_RE, page)
+ # We remove the duplicates and the link with index 0
+ # (it's not the first video of the playlist)
+ new_ids = orderedSet(m.group('id') for m in matches if m.group('index') != '0')
+ ids.extend(new_ids)
- try:
- response = json.loads(page)
- except ValueError as err:
- raise ExtractorError(u'Invalid JSON in API response: ' + compat_str(err))
-
- if 'feed' not in response:
- raise ExtractorError(u'Got a malformed response from YouTube API')
- playlist_title = response['feed']['title']['$t']
- if 'entry' not in response['feed']:
- # Number of videos is a multiple of self._MAX_RESULTS
+ if re.search(self._MORE_PAGES_INDICATOR, page) is None:
break
- for entry in response['feed']['entry']:
- index = entry['yt$position']['$t']
- if 'media$group' in entry and 'yt$videoid' in entry['media$group']:
- videos.append((
- index,
- 'https://www.youtube.com/watch?v=' + entry['media$group']['yt$videoid']['$t']
- ))
-
- videos = [v[1] for v in sorted(videos)]
+ playlist_title = self._og_search_title(page)
- url_results = [self.url_result(vurl, 'Youtube') for vurl in videos]
- return [self.playlist_result(url_results, playlist_id, playlist_title)]
+ url_results = self._ids_to_results(ids)
+ return self.playlist_result(url_results, playlist_id, playlist_title)
class YoutubeChannelIE(InfoExtractor):
self._downloader.to_screen(u'[youtube] Channel %s: Found %i videos' % (channel_id, len(video_ids)))
- urls = ['http://www.youtube.com/watch?v=%s' % id for id in video_ids]
- url_entries = [self.url_result(eurl, 'Youtube') for eurl in urls]
- return [self.playlist_result(url_entries, channel_id)]
+ url_entries = [self.url_result(video_id, 'Youtube', video_id=video_id)
+ for video_id in video_ids]
+ return self.playlist_result(url_entries, channel_id)
class YoutubeUserIE(InfoExtractor):
if len(ids_in_page) < self._GDATA_PAGE_SIZE:
break
- urls = ['http://www.youtube.com/watch?v=%s' % video_id for video_id in video_ids]
- url_results = [self.url_result(rurl, 'Youtube') for rurl in urls]
- return [self.playlist_result(url_results, playlist_title = username)]
+ url_results = [
+ self.url_result(video_id, 'Youtube', video_id=video_id)
+ for video_id in video_ids]
+ return self.playlist_result(url_results, playlist_title=username)
+
class YoutubeSearchIE(SearchInfoExtractor):
IE_DESC = u'YouTube.com searches'
if len(video_ids) > n:
video_ids = video_ids[:n]
- videos = [self.url_result('http://www.youtube.com/watch?v=%s' % id, 'Youtube') for id in video_ids]
+ videos = [self.url_result(video_id, 'Youtube', video_id=video_id)
+ for video_id in video_ids]
return self.playlist_result(videos, query)
class YoutubeSearchDateIE(YoutubeSearchIE):
+ IE_NAME = YoutubeSearchIE.IE_NAME + ':date'
_API_URL = 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc&orderby=published'
_SEARCH_KEY = 'ytsearchdate'
IE_DESC = u'YouTube.com searches, newest videos first'
Subclasses must define the _FEED_NAME and _PLAYLIST_TITLE properties.
"""
_LOGIN_REQUIRED = True
- _PAGING_STEP = 30
# use action_load_personal_feed instead of action_load_system_feed
_PERSONAL_FEED = False
def _real_extract(self, url):
feed_entries = []
- # The step argument is available only in 2.7 or higher
- for i in itertools.count(0):
- paging = i*self._PAGING_STEP
+ paging = 0
+ for i in itertools.count(1):
info = self._download_webpage(self._FEED_TEMPLATE % paging,
u'%s feed' % self._FEED_NAME,
u'Downloading page %s' % i)
feed_html = info['feed_html']
m_ids = re.finditer(r'"/watch\?v=(.*?)["&]', feed_html)
ids = orderedSet(m.group(1) for m in m_ids)
- feed_entries.extend(self.url_result(id, 'Youtube') for id in ids)
+ feed_entries.extend(
+ self.url_result(video_id, 'Youtube', video_id=video_id)
+ for video_id in ids)
if info['paging'] is None:
break
+ paging = info['paging']
return self.playlist_result(feed_entries, playlist_title=self._PLAYLIST_TITLE)
class YoutubeSubscriptionsIE(YoutubeFeedsInfoExtractor):
_VALID_URL = r'https?://www\.youtube\.com/feed/watch_later|:ytwatchlater'
_FEED_NAME = 'watch_later'
_PLAYLIST_TITLE = u'Youtube Watch Later'
- _PAGING_STEP = 100
_PERSONAL_FEED = True
+class YoutubeHistoryIE(YoutubeFeedsInfoExtractor):
+ IE_DESC = u'Youtube watch history, "ythistory" keyword (requires authentication)'
+ _VALID_URL = u'https?://www\.youtube\.com/feed/history|:ythistory'
+ _FEED_NAME = 'history'
+ _PERSONAL_FEED = True
+ _PLAYLIST_TITLE = u'Youtube Watch History'
+
class YoutubeFavouritesIE(YoutubeBaseInfoExtractor):
IE_NAME = u'youtube:favorites'
IE_DESC = u'YouTube.com favourite videos, "ytfav" keyword (requires authentication)'
projects / youtube-dl / blobdiff