diff options
| author | Remita Amine <remitamine@gmail.com> | 2021-02-01 14:30:59 +0100 |
|---|---|---|
| committer | Remita Amine <remitamine@gmail.com> | 2021-02-01 14:53:01 +0100 |
| commit | 67299f23d8b1894120e875edf97440de87e22308 (patch) | |
| tree | e5bd5dc3e3018d8ef99e83304a1d4a799fbb272a | |
| parent | 8bf9591a70757c624a8ea5bf686040ed752246e0 (diff) | |
| download | youtube-dl-67299f23d8b1894120e875edf97440de87e22308.tar.xz | |
[youtube] Rewrite Extractor
- improve format sorting
- remove unused code(swf parsing, ...)
- fix series metadata extraction
- fix trailer video extraction
- improve error reporting
- extract video location
| -rw-r--r-- | youtube_dl/extractor/common.py | 17 | ||||
| -rw-r--r-- | youtube_dl/extractor/youtube.py | 1735 |
2 files changed, 492 insertions, 1260 deletions
diff --git a/youtube_dl/extractor/common.py b/youtube_dl/extractor/common.py index d5faa0eb7..8eb110f4e 100644 --- a/youtube_dl/extractor/common.py +++ b/youtube_dl/extractor/common.py @@ -2064,7 +2064,7 @@ class InfoExtractor(object): }) return entries - def _extract_mpd_formats(self, mpd_url, video_id, mpd_id=None, note=None, errnote=None, fatal=True, formats_dict={}, data=None, headers={}, query={}): + def _extract_mpd_formats(self, mpd_url, video_id, mpd_id=None, note=None, errnote=None, fatal=True, data=None, headers={}, query={}): res = self._download_xml_handle( mpd_url, video_id, note=note or 'Downloading MPD manifest', @@ -2078,10 +2078,9 @@ class InfoExtractor(object): mpd_base_url = base_url(urlh.geturl()) return self._parse_mpd_formats( - mpd_doc, mpd_id=mpd_id, mpd_base_url=mpd_base_url, - formats_dict=formats_dict, mpd_url=mpd_url) + mpd_doc, mpd_id, mpd_base_url, mpd_url) - def _parse_mpd_formats(self, mpd_doc, mpd_id=None, mpd_base_url='', formats_dict={}, mpd_url=None): + def _parse_mpd_formats(self, mpd_doc, mpd_id=None, mpd_base_url='', mpd_url=None): """ Parse formats from MPD manifest. References: @@ -2359,15 +2358,7 @@ class InfoExtractor(object): else: # Assuming direct URL to unfragmented media. f['url'] = base_url - - # According to [1, 5.3.5.2, Table 7, page 35] @id of Representation - # is not necessarily unique within a Period thus formats with - # the same `format_id` are quite possible. There are numerous examples - # of such manifests (see https://github.com/ytdl-org/youtube-dl/issues/15111, - # https://github.com/ytdl-org/youtube-dl/issues/13919) - full_info = formats_dict.get(representation_id, {}).copy() - full_info.update(f) - formats.append(full_info) + formats.append(f) else: self.report_warning('Unknown MIME type %s in DASH manifest' % mime_type) return formats diff --git a/youtube_dl/extractor/youtube.py b/youtube_dl/extractor/youtube.py index b254ceced..5f6769878 100644 --- a/youtube_dl/extractor/youtube.py +++ b/youtube_dl/extractor/youtube.py @@ -2,41 +2,34 @@ from __future__ import unicode_literals - import itertools import json import os.path import random import re -import time import traceback from .common import InfoExtractor, SearchInfoExtractor -from ..jsinterp import JSInterpreter -from ..swfinterp import SWFInterpreter from ..compat import ( compat_chr, compat_HTTPError, compat_parse_qs, - compat_urllib_parse_unquote, + compat_str, compat_urllib_parse_unquote_plus, compat_urllib_parse_urlencode, compat_urllib_parse_urlparse, compat_urlparse, - compat_str, ) +from ..jsinterp import JSInterpreter from ..utils import ( - bool_or_none, - clean_html, - error_to_compat_str, ExtractorError, + clean_html, float_or_none, - get_element_by_id, int_or_none, mimetype2ext, parse_codecs, parse_duration, - remove_quotes, + qualities, remove_start, smuggle_url, str_or_none, @@ -46,7 +39,6 @@ from ..utils import ( unified_strdate, unsmuggle_url, update_url_query, - uppercase_escape, url_or_none, urlencode_postdata, urljoin, @@ -68,12 +60,6 @@ class YoutubeBaseInfoExtractor(InfoExtractor): _PLAYLIST_ID_RE = r'(?:(?:PL|LL|EC|UU|FL|RD|UL|TL|PU|OLAK5uy_)[0-9A-Za-z-_]{10,}|RDMM)' - def _set_language(self): - self._set_cookie( - '.youtube.com', 'PREF', 'f1=50000000&f6=8&hl=en', - # YouTube sets the expire time to about two months - expire_time=time.time() + 2 * 30 * 24 * 3600) - def _ids_to_results(self, ids): return [ self.url_result(vid_id, 'Youtube', video_id=vid_id) @@ -265,7 +251,6 @@ class YoutubeBaseInfoExtractor(InfoExtractor): def _real_initialize(self): if self._downloader is None: return - self._set_language() if not self._login(): return @@ -282,19 +267,17 @@ class YoutubeBaseInfoExtractor(InfoExtractor): _YT_INITIAL_PLAYER_RESPONSE_RE = r'ytInitialPlayerResponse\s*=\s*({.+?})\s*;' _YT_INITIAL_BOUNDARY_RE = r'(?:var\s+meta|</script|\n)' - def _call_api(self, ep, query, video_id): + def _call_api(self, ep, query, video_id, fatal=True): data = self._DEFAULT_API_DATA.copy() data.update(query) - response = self._download_json( + return self._download_json( 'https://www.youtube.com/youtubei/v1/%s' % ep, video_id=video_id, note='Downloading API JSON', errnote='Unable to download API page', - data=json.dumps(data).encode('utf8'), + data=json.dumps(data).encode('utf8'), fatal=fatal, headers={'content-type': 'application/json'}, query={'key': 'AIzaSyAO_FJ2SlqU8Q4STEHLGCilw_Y9_11qcW8'}) - return response - def _extract_yt_initial_data(self, video_id, webpage): return self._parse_json( self._search_regex( @@ -414,116 +397,10 @@ class YoutubeIE(YoutubeBaseInfoExtractor): ) (?(1).+)? # if we found the ID, everything can follow $""" % {'playlist_id': YoutubeBaseInfoExtractor._PLAYLIST_ID_RE} - _NEXT_URL_RE = r'[\?&]next_url=([^&]+)' _PLAYER_INFO_RE = ( - r'/(?P<id>[a-zA-Z0-9_-]{8,})/player_ias\.vflset(?:/[a-zA-Z]{2,3}_[a-zA-Z]{2,3})?/base\.(?P<ext>[a-z]+)$', - r'\b(?P<id>vfl[a-zA-Z0-9_-]+)\b.*?\.(?P<ext>[a-z]+)$', + r'/(?P<id>[a-zA-Z0-9_-]{8,})/player_ias\.vflset(?:/[a-zA-Z]{2,3}_[a-zA-Z]{2,3})?/base\.js$', + r'\b(?P<id>vfl[a-zA-Z0-9_-]+)\b.*?\.js$', ) - _formats = { - '5': {'ext': 'flv', 'width': 400, 'height': 240, 'acodec': 'mp3', 'abr': 64, 'vcodec': 'h263'}, - '6': {'ext': 'flv', 'width': 450, 'height': 270, 'acodec': 'mp3', 'abr': 64, 'vcodec': 'h263'}, - '13': {'ext': '3gp', 'acodec': 'aac', 'vcodec': 'mp4v'}, - '17': {'ext': '3gp', 'width': 176, 'height': 144, 'acodec': 'aac', 'abr': 24, 'vcodec': 'mp4v'}, - '18': {'ext': 'mp4', 'width': 640, 'height': 360, 'acodec': 'aac', 'abr': 96, 'vcodec': 'h264'}, - '22': {'ext': 'mp4', 'width': 1280, 'height': 720, 'acodec': 'aac', 'abr': 192, 'vcodec': 'h264'}, - '34': {'ext': 'flv', 'width': 640, 'height': 360, 'acodec': 'aac', 'abr': 128, 'vcodec': 'h264'}, - '35': {'ext': 'flv', 'width': 854, 'height': 480, 'acodec': 'aac', 'abr': 128, 'vcodec': 'h264'}, - # itag 36 videos are either 320x180 (BaW_jenozKc) or 320x240 (__2ABJjxzNo), abr varies as well - '36': {'ext': '3gp', 'width': 320, 'acodec': 'aac', 'vcodec': 'mp4v'}, - '37': {'ext': 'mp4', 'width': 1920, 'height': 1080, 'acodec': 'aac', 'abr': 192, 'vcodec': 'h264'}, - '38': {'ext': 'mp4', 'width': 4096, 'height': 3072, 'acodec': 'aac', 'abr': 192, 'vcodec': 'h264'}, - '43': {'ext': 'webm', 'width': 640, 'height': 360, 'acodec': 'vorbis', 'abr': 128, 'vcodec': 'vp8'}, - '44': {'ext': 'webm', 'width': 854, 'height': 480, 'acodec': 'vorbis', 'abr': 128, 'vcodec': 'vp8'}, - '45': {'ext': 'webm', 'width': 1280, 'height': 720, 'acodec': 'vorbis', 'abr': 192, 'vcodec': 'vp8'}, - '46': {'ext': 'webm', 'width': 1920, 'height': 1080, 'acodec': 'vorbis', 'abr': 192, 'vcodec': 'vp8'}, - '59': {'ext': 'mp4', 'width': 854, 'height': 480, 'acodec': 'aac', 'abr': 128, 'vcodec': 'h264'}, - '78': {'ext': 'mp4', 'width': 854, 'height': 480, 'acodec': 'aac', 'abr': 128, 'vcodec': 'h264'}, - - - # 3D videos - '82': {'ext': 'mp4', 'height': 360, 'format_note': '3D', 'acodec': 'aac', 'abr': 128, 'vcodec': 'h264', 'preference': -20}, - '83': {'ext': 'mp4', 'height': 480, 'format_note': '3D', 'acodec': 'aac', 'abr': 128, 'vcodec': 'h264', 'preference': -20}, - '84': {'ext': 'mp4', 'height': 720, 'format_note': '3D', 'acodec': 'aac', 'abr': 192, 'vcodec': 'h264', 'preference': -20}, - '85': {'ext': 'mp4', 'height': 1080, 'format_note': '3D', 'acodec': 'aac', 'abr': 192, 'vcodec': 'h264', 'preference': -20}, - '100': {'ext': 'webm', 'height': 360, 'format_note': '3D', 'acodec': 'vorbis', 'abr': 128, 'vcodec': 'vp8', 'preference': -20}, - '101': {'ext': 'webm', 'height': 480, 'format_note': '3D', 'acodec': 'vorbis', 'abr': 192, 'vcodec': 'vp8', 'preference': -20}, - '102': {'ext': 'webm', 'height': 720, 'format_note': '3D', 'acodec': 'vorbis', 'abr': 192, 'vcodec': 'vp8', 'preference': -20}, - - # Apple HTTP Live Streaming - '91': {'ext': 'mp4', 'height': 144, 'format_note': 'HLS', 'acodec': 'aac', 'abr': 48, 'vcodec': 'h264', 'preference': -10}, - '92': {'ext': 'mp4', 'height': 240, 'format_note': 'HLS', 'acodec': 'aac', 'abr': 48, 'vcodec': 'h264', 'preference': -10}, - '93': {'ext': 'mp4', 'height': 360, 'format_note': 'HLS', 'acodec': 'aac', 'abr': 128, 'vcodec': 'h264', 'preference': -10}, - '94': {'ext': 'mp4', 'height': 480, 'format_note': 'HLS', 'acodec': 'aac', 'abr': 128, 'vcodec': 'h264', 'preference': -10}, - '95': {'ext': 'mp4', 'height': 720, 'format_note': 'HLS', 'acodec': 'aac', 'abr': 256, 'vcodec': 'h264', 'preference': -10}, - '96': {'ext': 'mp4', 'height': 1080, 'format_note': 'HLS', 'acodec': 'aac', 'abr': 256, 'vcodec': 'h264', 'preference': -10}, - '132': {'ext': 'mp4', 'height': 240, 'format_note': 'HLS', 'acodec': 'aac', 'abr': 48, 'vcodec': 'h264', 'preference': -10}, - '151': {'ext': 'mp4', 'height': 72, 'format_note': 'HLS', 'acodec': 'aac', 'abr': 24, 'vcodec': 'h264', 'preference': -10}, - - # DASH mp4 video - '133': {'ext': 'mp4', 'height': 240, 'format_note': 'DASH video', 'vcodec': 'h264'}, - '134': {'ext': 'mp4', 'height': 360, 'format_note': 'DASH video', 'vcodec': 'h264'}, - '135': {'ext': 'mp4', 'height': 480, 'format_note': 'DASH video', 'vcodec': 'h264'}, - '136': {'ext': 'mp4', 'height': 720, 'format_note': 'DASH video', 'vcodec': 'h264'}, - '137': {'ext': 'mp4', 'height': 1080, 'format_note': 'DASH video', 'vcodec': 'h264'}, - '138': {'ext': 'mp4', 'format_note': 'DASH video', 'vcodec': 'h264'}, # Height can vary (https://github.com/ytdl-org/youtube-dl/issues/4559) - '160': {'ext': 'mp4', 'height': 144, 'format_note': 'DASH video', 'vcodec': 'h264'}, - '212': {'ext': 'mp4', 'height': 480, 'format_note': 'DASH video', 'vcodec': 'h264'}, - '264': {'ext': 'mp4', 'height': 1440, 'format_note': 'DASH video', 'vcodec': 'h264'}, - '298': {'ext': 'mp4', 'height': 720, 'format_note': 'DASH video', 'vcodec': 'h264', 'fps': 60}, - '299': {'ext': 'mp4', 'height': 1080, 'format_note': 'DASH video', 'vcodec': 'h264', 'fps': 60}, - '266': {'ext': 'mp4', 'height': 2160, 'format_note': 'DASH video', 'vcodec': 'h264'}, - - # Dash mp4 audio - '139': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'aac', 'abr': 48, 'container': 'm4a_dash'}, - '140': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'aac', 'abr': 128, 'container': 'm4a_dash'}, - '141': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'aac', 'abr': 256, 'container': 'm4a_dash'}, - '256': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'aac', 'container': 'm4a_dash'}, - '258': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'aac', 'container': 'm4a_dash'}, - '325': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'dtse', 'container': 'm4a_dash'}, - '328': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'ec-3', 'container': 'm4a_dash'}, - - # Dash webm - '167': {'ext': 'webm', 'height': 360, 'width': 640, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'vp8'}, - '168': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'vp8'}, - '169': {'ext': 'webm', 'height': 720, 'width': 1280, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'vp8'}, - '170': {'ext': 'webm', 'height': 1080, 'width': 1920, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'vp8'}, - '218': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'vp8'}, - '219': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'vp8'}, - '278': {'ext': 'webm', 'height': 144, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'vp9'}, - '242': {'ext': 'webm', 'height': 240, 'format_note': 'DASH video', 'vcodec': 'vp9'}, - '243': {'ext': 'webm', 'height': 360, 'format_note': 'DASH video', 'vcodec': 'vp9'}, - '244': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'vcodec': 'vp9'}, - '245': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'vcodec': 'vp9'}, - '246': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'vcodec': 'vp9'}, - '247': {'ext': 'webm', 'height': 720, 'format_note': 'DASH video', 'vcodec': 'vp9'}, - '248': {'ext': 'webm', 'height': 1080, 'format_note': 'DASH video', 'vcodec': 'vp9'}, - '271': {'ext': 'webm', 'height': 1440, 'format_note': 'DASH video', 'vcodec': 'vp9'}, - # itag 272 videos are either 3840x2160 (e.g. RtoitU2A-3E) or 7680x4320 (sLprVF6d7Ug) - '272': {'ext': 'webm', 'height': 2160, 'format_note': 'DASH video', 'vcodec': 'vp9'}, - '302': {'ext': 'webm', 'height': 720, 'format_note': 'DASH video', 'vcodec': 'vp9', 'fps': 60}, - '303': {'ext': 'webm', 'height': 1080, 'format_note': 'DASH video', 'vcodec': 'vp9', 'fps': 60}, - '308': {'ext': 'webm', 'height': 1440, 'format_note': 'DASH video', 'vcodec': 'vp9', 'fps': 60}, - '313': {'ext': 'webm', 'height': 2160, 'format_note': 'DASH video', 'vcodec': 'vp9'}, - '315': {'ext': 'webm', 'height': 2160, 'format_note': 'DASH video', 'vcodec': 'vp9', 'fps': 60}, - - # Dash webm audio - '171': {'ext': 'webm', 'acodec': 'vorbis', 'format_note': 'DASH audio', 'abr': 128}, - '172': {'ext': 'webm', 'acodec': 'vorbis', 'format_note': 'DASH audio', 'abr': 256}, - - # Dash webm audio with opus inside - '249': {'ext': 'webm', 'format_note': 'DASH audio', 'acodec': 'opus', 'abr': 50}, - '250': {'ext': 'webm', 'format_note': 'DASH audio', 'acodec': 'opus', 'abr': 70}, - '251': {'ext': 'webm', 'format_note': 'DASH audio', 'acodec': 'opus', 'abr': 160}, - - # RTMP (unnamed) - '_rtmp': {'protocol': 'rtmp'}, - - # av01 video only formats sometimes served with "unknown" codecs - '394': {'acodec': 'none', 'vcodec': 'av01.0.05M.08'}, - '395': {'acodec': 'none', 'vcodec': 'av01.0.05M.08'}, - '396': {'acodec': 'none', 'vcodec': 'av01.0.05M.08'}, - '397': {'acodec': 'none', 'vcodec': 'av01.0.05M.08'}, - } _SUBTITLE_FORMATS = ('srv1', 'srv2', 'srv3', 'ttml', 'vtt') _GEO_BYPASS = False @@ -566,7 +443,8 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 'uploader_id': 'setindia', 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/setindia', 'age_limit': 18, - } + }, + 'skip': 'Private video', }, { 'url': 'https://www.youtube.com/watch?v=BaW_jenozKc&v=yZIXLfi8CZQ', @@ -640,7 +518,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 'uploader_id': 'TheAmazingAtheist', 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/TheAmazingAtheist', 'title': 'Burning Everyone\'s Koran', - 'description': 'SUBSCRIBE: http://www.youtube.com/saturninefilms\n\nEven Obama has taken a stand against freedom on this issue: http://www.huffingtonpost.com/2010/09/09/obama-gma-interview-quran_n_710282.html', + 'description': 'SUBSCRIBE: http://www.youtube.com/saturninefilms \r\n\r\nEven Obama has taken a stand against freedom on this issue: http://www.huffingtonpost.com/2010/09/09/obama-gma-interview-quran_n_710282.html', } }, # Normal age-gate video (No vevo, embed allowed), available via embed page @@ -676,11 +554,11 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 'upload_date': '20100430', 'uploader_id': 'deadmau5', 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/deadmau5', - 'creator': 'Dada Life, deadmau5', - 'description': 'md5:12c56784b8032162bb936a5f76d55360', + 'creator': 'deadmau5', + 'description': 'md5:6cbcd3a92ce1bc676fc4d6ab4ace2336', 'uploader': 'deadmau5', 'title': 'Deadmau5 - Some Chords (HD)', - 'alt_title': 'This Machine Kills Some Chords', + 'alt_title': 'Some Chords', }, 'expected_warnings': [ 'DASH manifest missing', @@ -775,69 +653,64 @@ class YoutubeIE(YoutubeBaseInfoExtractor): }, { # Multifeed videos (multiple cameras), URL is for Main Camera - 'url': 'https://www.youtube.com/watch?v=jqWvoWXjCVs', + 'url': 'https://www.youtube.com/watch?v=jvGDaLqkpTg', 'info_dict': { - 'id': 'jqWvoWXjCVs', - 'title': 'teamPGP: Rocket League Noob Stream', - 'description': 'md5:dc7872fb300e143831327f1bae3af010', + 'id': 'jvGDaLqkpTg', + 'title': 'Tom Clancy Free Weekend Rainbow Whatever', + 'description': 'md5:e03b909557865076822aa169218d6a5d', }, 'playlist': [{ 'info_dict': { - 'id': 'jqWvoWXjCVs', + 'id': 'jvGDaLqkpTg', 'ext': 'mp4', - 'title': 'teamPGP: Rocket League Noob Stream (Main Camera)', - 'description': 'md5:dc7872fb300e143831327f1bae3af010', - 'duration': 7335, - 'upload_date': '20150721', - 'uploader': 'Beer Games Beer', - 'uploader_id': 'beergamesbeer', - 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/beergamesbeer', - 'license': 'Standard YouTube License', + 'title': 'Tom Clancy Free Weekend Rainbow Whatever (Main Camera)', + 'description': 'md5:e03b909557865076822aa169218d6a5d', + 'duration': 10643, + 'upload_date': '20161111', + 'uploader': 'Team PGP', + 'uploader_id': 'UChORY56LMMETTuGjXaJXvLg', + 'uploader_url': r're:https?://(?:www\.)?youtube\.com/channel/UChORY56LMMETTuGjXaJXvLg', }, }, { 'info_dict': { - 'id': '6h8e8xoXJzg', + 'id': '3AKt1R1aDnw', 'ext': 'mp4', - 'title': 'teamPGP: Rocket League Noob Stream (kreestuh)', - 'description': 'md5:dc7872fb300e143831327f1bae3af010', - 'duration': 7337, - 'upload_date': '20150721', - 'uploader': 'Beer Games Beer', - 'uploader_id': 'beergamesbeer', - 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/beergamesbeer', - 'license': 'Standard YouTube License', + 'title': 'Tom Clancy Free Weekend Rainbow Whatever (Camera 2)', + 'description': 'md5:e03b909557865076822aa169218d6a5d', + 'duration': 10991, + 'upload_date': '20161111', + 'uploader': 'Team PGP', + 'uploader_id': 'UChORY56LMMETTuGjXaJXvLg', + 'uploader_url': r're:https?://(?:www\.)?youtube\.com/channel/UChORY56LMMETTuGjXaJXvLg', }, }, { 'info_dict': { - 'id': 'PUOgX5z9xZw', + 'id': 'RtAMM00gpVc', 'ext': 'mp4', - 'title': 'teamPGP: Rocket League Noob Stream (grizzle)', - 'description': 'md5:dc7872fb300e143831327f1bae3af010', - 'duration': 7337, - 'upload_date': '20150721', - 'uploader': 'Beer Games Beer', - 'uploader_id': 'beergamesbeer', - 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/beergamesbeer', - 'license': 'Standard YouTube License', + 'title': 'Tom Clancy Free Weekend Rainbow Whatever (Camera 3)', + 'description': 'md5:e03b909557865076822aa169218d6a5d', + 'duration': 10995, + 'upload_date': '20161111', + 'uploader': 'Team PGP', + 'uploader_id': 'UChORY56LMMETTuGjXaJXvLg', + 'uploader_url': r're:https?://(?:www\.)?youtube\.com/channel/UChORY56LMMETTuGjXaJXvLg', }, }, { 'info_dict': { - 'id': 'teuwxikvS5k', + 'id': '6N2fdlP3C5U', 'ext': 'mp4', - 'title': 'teamPGP: Rocket League Noob Stream (zim)', - 'description': 'md5:dc7872fb300e143831327f1bae3af010', - 'duration': 7334, - 'upload_date': '20150721', - 'uploader': 'Beer Games Beer', - 'uploader_id': 'beergamesbeer', - 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/beergamesbeer', - 'license': 'Standard YouTube License', + 'title': 'Tom Clancy Free Weekend Rainbow Whatever (Camera 4)', + 'description': 'md5:e03b909557865076822aa169218d6a5d', + 'duration': 10990, + 'upload_date': '20161111', + 'uploader': 'Team PGP', + 'uploader_id': 'UChORY56LMMETTuGjXaJXvLg', + 'uploader_url': r're:https?://(?:www\.)?youtube\.com/channel/UChORY56LMMETTuGjXaJXvLg', }, }], 'params': { 'skip_download': True, }, - 'skip': 'This video is not available.', }, { # Multifeed video with comma in title (see https://github.com/ytdl-org/youtube-dl/issues/8536) @@ -931,7 +804,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 'id': 'eQcmzGIKrzg', 'ext': 'mp4', 'title': 'Democratic Socialism and Foreign Policy | Bernie Sanders', - 'description': 'md5:dda0d780d5a6e120758d1711d062a867', + 'description': 'md5:13a2503d7b5904ef4b223aa101628f39', 'duration': 4060, 'upload_date': '20151119', 'uploader': 'Bernie Sanders', @@ -978,7 +851,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 'id': 'iqKdEhx-dD4', 'ext': 'mp4', 'title': 'Isolation - Mind Field (Ep 1)', - 'description': 'md5:46a29be4ceffa65b92d277b93f463c0f', + 'description': 'md5:f540112edec5d09fc8cc752d3d4ba3cd', 'duration': 2085, 'upload_date': '20170118', 'uploader': 'Vsauce', @@ -1013,6 +886,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 'params': { 'skip_download': True, }, + 'skip': 'This video has been removed for violating YouTube\'s policy on hate speech.', }, { # itag 212 @@ -1098,6 +972,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 'params': { 'skip_download': True, }, + 'skip': 'Video unavailable', }, { # empty description results in an empty string @@ -1147,24 +1022,9 @@ class YoutubeIE(YoutubeBaseInfoExtractor): def __init__(self, *args, **kwargs): super(YoutubeIE, self).__init__(*args, **kwargs) + self._code_cache = {} self._player_cache = {} - def report_video_info_webpage_download(self, video_id): - """Report attempt to download video info webpage.""" - self.to_screen('%s: Downloading video info webpage' % video_id) - - def report_information_extraction(self, video_id): - """Report attempt to extract video information.""" - self.to_screen('%s: Extracting video information' % video_id) - - def report_unavailable_format(self, video_id, format): - """Report extracted video URL.""" - self.to_screen('%s: Format %s not available' % (video_id, format)) - - def report_rtmp_download(self): - """Indicate the download will use the RTMP protocol.""" - self.to_screen('RTMP download detected') - def _signature_cache_id(self, example_sig): """ Return a string representation of a signature """ return '.'.join(compat_str(len(part)) for part in example_sig.split('.')) @@ -1177,40 +1037,27 @@ class YoutubeIE(YoutubeBaseInfoExtractor): break else: raise ExtractorError('Cannot identify player %r' % player_url) - return id_m.group('ext'), id_m.group('id') + return id_m.group('id') def _extract_signature_function(self, video_id, player_url, example_sig): - player_type, player_id = self._extract_player_info(player_url) + player_id = self._extract_player_info(player_url) # Read from filesystem cache - func_id = '%s_%s_%s' % ( - player_type, player_id, self._signature_cache_id(example_sig)) + func_id = 'js_%s_%s' % ( + player_id, self._signature_cache_id(example_sig)) assert os.path.basename(func_id) == func_id cache_spec = self._downloader.cache.load('youtube-sigfuncs', func_id) if cache_spec is not None: return lambda s: ''.join(s[i] for i in cache_spec) - download_note = ( - 'Downloading player %s' % player_url - if self._downloader.params.get('verbose') else - 'Downloading %s player %s' % (player_type, player_id) - ) - if player_type == 'js': - code = self._download_webpage( - player_url, video_id, - note=download_note, - errnote='Download of %s failed' % player_url) - res = self._parse_sig_js(code) - elif player_type == 'swf': - urlh = self._request_webpage( + if player_id not in self._code_cache: + self._code_cache[player_id] = self._download_webpage( player_url, video_id, - note=download_note, + note='Downloading player ' + player_id, errnote='Download of %s failed' % player_url) - code = urlh.read() - res = self._parse_sig_swf(code) - else: - assert False, 'Invalid player type %r' % player_type + code = self._code_cache[player_id] + res = self._parse_sig_js(code) test_string = ''.join(map(compat_chr, range(len(example_sig)))) cache_res = res(test_string) @@ -1279,14 +1126,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor): initial_function = jsi.extract_function(funcname) return lambda s: initial_function([s]) - def _parse_sig_swf(self, file_contents): - swfi = SWFInterpreter(file_contents) - TARGET_CLASSNAME = 'SignatureDecipher' - searched_class = swfi.extract_class(TARGET_CLASSNAME) - initial_function = swfi.extract_function(searched_class, 'decipher') - return lambda s: initial_function([s]) - - def _decrypt_signature(self, s, video_id, player_url, age_gate=False): + def _decrypt_signature(self, s, video_id, player_url): """Turn the encrypted s field into a working signature""" if player_url is None: @@ -1313,158 +1153,10 @@ class YoutubeIE(YoutubeBaseInfoExtractor): raise ExtractorError( 'Signature extraction failed: ' + tb, cause=e) - def _get_subtitles(self, video_id, webpage): - try: - subs_doc = self._download_xml( - 'https://video.google.com/timedtext?hl=en&type=list&v=%s' % video_id, - video_id, note=False) - except ExtractorError as err: - self._downloader.report_warning('unable to download video subtitles: %s' % error_to_compat_str(err)) - return {} - - sub_lang_list = {} - for track in subs_doc.findall('track'): - lang = track.attrib['lang_code'] - if lang in sub_lang_list: - continue - sub_formats = [] - for ext in self._SUBTITLE_FORMATS: - params = compat_urllib_parse_urlencode({ - 'lang': lang, - 'v': video_id, - 'fmt': ext, - 'name': track.attrib['name'].encode('utf-8'), - }) - sub_formats.append({ - 'url': 'https://www.youtube.com/api/timedtext?' + params, - 'ext': ext, - }) - sub_lang_list[lang] = sub_formats - if not sub_lang_list: - self._downloader.report_warning('video doesn\'t have subtitles') - return {} - return sub_lang_list - - def _get_ytplayer_config(self, video_id, webpage): - patterns = ( - # User data may contain arbitrary character sequences that may affect - # JSON extraction with regex, e.g. when '};' is contained the second - # regex won't capture the whole JSON. Yet working around by trying more - # concrete regex first keeping in mind proper quoted string handling - # to be implemented in future that will replace this workaround (see - # https://github.com/ytdl-org/youtube-dl/issues/7468, - # https://github.com/ytdl-org/youtube-dl/pull/7599) - r';ytplayer\.config\s*=\s*({.+?});ytplayer', - r';ytplayer\.config\s*=\s*({.+?});', - ) - config = self._search_regex( - patterns, webpage, 'ytplayer.config', default=None) - if config: - return self._parse_json( - uppercase_escape(config), video_id, fatal=False) - - def _get_automatic_captions(self, video_id, player_response, player_config): - """We need the webpage for getting the captions url, pass it as an - argument to speed up the process.""" - self.to_screen('%s: Looking for automatic captions' % video_id) - err_msg = 'Couldn\'t find automatic captions for %s' % video_id - if not (player_response or player_config): - self._downloader.report_warning(err_msg) - return {} - try: - args = player_config.get('args') if player_config else {} - caption_url = args.get('ttsurl') - if caption_url: - timestamp = args['timestamp'] - # We get the available subtitles - list_params = compat_urllib_parse_urlencode({ - 'type': 'list', - 'tlangs': 1, - 'asrs': 1, - }) - list_url = caption_url + '&' + list_params - caption_list = self._download_xml(list_url, video_id) - original_lang_node = caption_list.find('track') - if original_lang_node is None: - self._downloader.report_warning('Video doesn\'t have automatic captions') - return {} - original_lang = original_lang_node.attrib['lang_code'] - caption_kind = original_lang_node.attrib.get('kind', '') - - sub_lang_list = {} - for lang_node in caption_list.findall('target'): - sub_lang = lang_node.attrib['lang_code'] - sub_formats = [] - for ext in self._SUBTITLE_FORMATS: - params = compat_urllib_parse_urlencode({ - 'lang': original_lang, - 'tlang': sub_lang, - 'fmt': ext, - 'ts': timestamp, - 'kind': caption_kind, - }) - sub_formats.append({ - 'url': caption_url + '&' + params, - 'ext': ext, - }) - 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 - 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] - 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 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, IndexError, ExtractorError): - self._downloader.report_warning(err_msg) - return {} - - def _mark_watched(self, video_id, video_info, player_response): + def _mark_watched(self, video_id, player_response): playback_url = url_or_none(try_get( player_response, - lambda x: x['playbackTracking']['videostatsPlaybackUrl']['baseUrl']) or try_get( - video_info, lambda x: x['videostats_playback_base_url'][0])) + lambda x: x['playbackTracking']['videostatsPlaybackUrl']['baseUrl'])) if not playback_url: return parsed_playback_url = compat_urlparse.urlparse(playback_url) @@ -1531,289 +1223,74 @@ class YoutubeIE(YoutubeBaseInfoExtractor): video_id = mobj.group(2) return video_id - def _extract_chapters_from_json(self, webpage, video_id, duration): - if not webpage: - return - data = self._extract_yt_initial_data(video_id, webpage) - if not data or not isinstance(data, dict): - return - chapters_list = try_get( - data, - lambda x: x['playerOverlays'] - ['playerOverlayRenderer'] - ['decoratedPlayerBarRenderer'] - ['decoratedPlayerBarRenderer'] - ['playerBar'] - ['chapteredPlayerBarRenderer'] - ['chapters'], - list) - if not chapters_list: - return - - def chapter_time(chapter): - return float_or_none( - try_get( - chapter, - lambda x: x['chapterRenderer']['timeRangeStartMillis'], - int), - scale=1000) - chapters = [] - for next_num, chapter in enumerate(chapters_list, start=1): - start_time = chapter_time(chapter) - if start_time is None: - continue - end_time = (chapter_time(chapters_list[next_num]) - if next_num < len(chapters_list) else duration) - if end_time is None: - continue - title = try_get( - chapter, lambda x: x['chapterRenderer']['title']['simpleText'], - compat_str) - chapters.append({ - 'start_time': start_time, - 'end_time': end_time, - 'title': title, - }) - return chapters - - @staticmethod - def _extract_chapters_from_description(description, duration): - if not description: - return None - chapter_lines = re.findall( - r'(?:^|<br\s*/>)([^<]*<a[^>]+onclick=["\']yt\.www\.watch\.player\.seekTo[^>]+>(\d{1,2}:\d{1,2}(?::\d{1,2})?)</a>[^>]*)(?=$|<br\s*/>)', - description) - if not chapter_lines: - return None - chapters = [] - for next_num, (chapter_line, time_point) in enumerate( - chapter_lines, start=1): - 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 - 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) - chapters.append({ - 'start_time': start_time, - 'end_time': end_time, - 'title': chapter_title, - }) - return chapters - - def _extract_chapters(self, webpage, description, video_id, duration): - return (self._extract_chapters_from_json(webpage, video_id, duration) - or self._extract_chapters_from_description(description, duration)) + def _extract_yt_initial_variable(self, webpage, regex, video_id, name): + return self._parse_json(self._search_regex( + (r'%s\s*%s' % (regex, self._YT_INITIAL_BOUNDARY_RE), + regex), webpage, name, default='{}'), video_id, fatal=False) def _real_extract(self, url): url, smuggled_data = unsmuggle_url(url, {}) + video_id = self._match_id(url) + base_url = self.http_scheme() + '//www.youtube.com/' + webpage_url = base_url + 'watch?v=' + video_id + webpage = self._download_webpage(webpage_url, video_id, fatal=False) + + player_response = None + if webpage: + player_response = self._extract_yt_initial_variable( + webpage, self._YT_INITIAL_PLAYER_RESPONSE_RE, + video_id, 'initial player response') + if not player_response: + player_response = self._call_api( + 'player', {'videoId': video_id}, video_id) + + playability_status = player_response.get('playabilityStatus') or {} + if playability_status.get('reason') == 'Sign in to confirm your age': + pr = self._parse_json(try_get(compat_parse_qs( + self._download_webpage( + base_url + 'get_video_info', video_id, + 'Refetching age-gated info webpage', + 'unable to download video info webpage', query={ + 'video_id': video_id, + }, fatal=False)), + lambda x: x['player_response'][0], + compat_str) or '{}', video_id) + if pr: + player_response = pr + + trailer_video_id = try_get( + playability_status, + lambda x: x['errorScreen']['playerLegacyDesktopYpcTrailerRenderer']['trailerVideoId'], + compat_str) + if trailer_video_id: + return self.url_result( + trailer_video_id, self.ie_key(), trailer_video_id) - proto = ( - 'http' if self._downloader.params.get('prefer_insecure', False) - else 'https') - - start_time = None - end_time = None - parsed_url = compat_urllib_parse_urlparse(url) - for component in [parsed_url.fragment, parsed_url.query]: - query = compat_parse_qs(component) - if start_time is None and 't' in query: - start_time = parse_duration(query['t'][0]) - if start_time is None and 'start' in query: - start_time = parse_duration(query['start'][0]) - if end_time is None and 'end' in query: - end_time = parse_duration(query['end'][0]) - - # Extract original video URL from URL with redirection, like age verification, using next_url parameter - mobj = re.search(self._NEXT_URL_RE, url) - if mobj: - url = proto + '://www.youtube.com/' + compat_urllib_parse_unquote(mobj.group(1)).lstrip('/') - video_id = self.extract_id(url) - - # Get video webpage - url = proto + '://www.youtube.com/watch?v=%s&gl=US&hl=en&has_verified=1&bpctr=9999999999' % video_id - video_webpage, urlh = self._download_webpage_handle(url, video_id) - - qs = compat_parse_qs(compat_urllib_parse_urlparse(urlh.geturl()).query) - video_id = qs.get('v', [None])[0] or video_id - - # Attempt to extract SWF player URL - mobj = re.search(r'swfConfig.*?"(https?:\\/\\/.*?watch.*?-.*?\.swf)"', video_webpage) - if mobj is not None: - player_url = re.sub(r'\\(.)', r'\1', mobj.group(1)) - else: - player_url = None - - dash_mpds = [] - - def add_dash_mpd(video_info): - dash_mpd = video_info.get('dashmpd') - if dash_mpd and dash_mpd[0] not in dash_mpds: - dash_mpds.append(dash_mpd[0]) - - def add_dash_mpd_pr(pl_response): - dash_mpd = url_or_none(try_get( - pl_response, lambda x: x['streamingData']['dashManifestUrl'], - compat_str)) - if dash_mpd and dash_mpd not in dash_mpds: - dash_mpds.append(dash_mpd) - - 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])) - - def extract_player_response(player_response, video_id): - pl_response = str_or_none(player_response) - if not pl_response: + def get_text(x): + if not x: return - pl_response = self._parse_json(pl_response, video_id, fatal=False) - if isinstance(pl_response, dict): - add_dash_mpd_pr(pl_response) - return pl_response - - player_response = {} - - # Get video info - video_info = {} - embed_webpage = None - ytplayer_config = None - - if re.search(r'["\']status["\']\s*:\s*["\']LOGIN_REQUIRED', video_webpage) is not None: - age_gate = True - # We simulate the access to the video from www.youtube.com/v/{video_id} - # this can be viewed without login into Youtube - url = proto + '://www.youtube.com/embed/%s' % video_id - embed_webpage = self._download_webpage(url, video_id, 'Downloading embed webpage') - data = compat_urllib_parse_urlencode({ - 'video_id': video_id, - 'eurl': 'https://youtube.googleapis.com/v/' + video_id, - 'sts': self._search_regex( - r'"sts"\s*:\s*(\d+)', embed_webpage, 'sts', default=''), - }) - video_info_url = proto + '://www.youtube.com/get_video_info?' + data - try: - video_info_webpage = self._download_webpage( - video_info_url, video_id, - note='Refetching age-gated info webpage', - errnote='unable to download video info webpage') - except ExtractorError: - video_info_webpage = None - if video_info_webpage: - video_info = compat_parse_qs(video_info_webpage) - pl_response = video_info.get('player_response', [None])[0] - player_response = extract_player_response(pl_response, video_id) - add_dash_mpd(video_info) - view_count = extract_view_count(video_info) - else: - age_gate = False - # Try looking directly into the video webpage - 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') 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) - # Rental video is not rented but preview is available (e.g. - # https://www.youtube.com/watch?v=yYr8q0y5Jfg, - # https://github.com/ytdl-org/youtube-dl/issues/10532) - if not video_info and args.get('ypc_vid'): - return self.url_result( - 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 - if not player_response: - player_response = extract_player_response(args.get('player_response'), video_id) - if not video_info or self._downloader.params.get('youtube_include_dash_manifest', True): - add_dash_mpd_pr(player_response) - - if not video_info and not player_response: - player_response = extract_player_response( - self._search_regex( - (r'%s\s*%s' % (self._YT_INITIAL_PLAYER_RESPONSE_RE, self._YT_INITIAL_BOUNDARY_RE), - self._YT_INITIAL_PLAYER_RESPONSE_RE), video_webpage, - 'initial player response', default='{}'), - video_id) - - def extract_unavailable_message(): - messages = [] - for tag, kind in (('h1', 'message'), ('div', 'submessage')): - msg = self._html_search_regex( - r'(?s)<{tag}[^>]+id=["\']unavailable-{kind}["\'][^>]*>(.+?)</{tag}>'.format(tag=tag, kind=kind), - video_webpage, 'unavailable %s' % kind, default=None) - if msg: - messages.append(msg) - if messages: - return '\n'.join(messages) - - if not video_info and not player_response: - unavailable_message = extract_unavailable_message() - if not unavailable_message: - unavailable_message = 'Unable to extract video data' - raise ExtractorError( - 'YouTube said: %s' % unavailable_message, expected=True, video_id=video_id) + return x.get('simpleText') or ''.join([r['text'] for r in x['runs']]) - if not isinstance(video_info, dict): - video_info = {} - - video_details = try_get( - player_response, lambda x: x['videoDetails'], dict) or {} + search_meta = ( + lambda x: self._html_search_meta(x, webpage, default=None)) \ + if webpage else lambda x: None + video_details = player_response.get('videoDetails') or {} microformat = try_get( - player_response, lambda x: x['microformat']['playerMicroformatRenderer'], dict) or {} - - video_title = video_info.get('title', [None])[0] or video_details.get('title') - if not video_title: - self._downloader.report_warning('Unable to extract video title') - video_title = '_' - - 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+)*? - (?:title|href)="([^"]+)"\s+ - (?:[a-zA-Z-]+="[^"]*"\s+)*? - class="[^"]*"[^>]*> - [^<]+\.{3}\s* - </a> - ''', replace_url, video_description) - video_description = clean_html(video_description) - else: - video_description = video_details.get('shortDescription') - if video_description is None: - video_description = self._html_search_meta('description', video_webpage) + player_response, + lambda x: x['microformat']['playerMicroformatRenderer'], + dict) or {} + video_title = video_details.get('title') \ + or get_text(microformat.get('title')) \ + or search_meta(['og:title', 'twitter:title', 'title']) + video_description = video_details.get('shortDescription') if not smuggled_data.get('force_singlefeed', False): if not self._downloader.params.get('noplaylist'): multifeed_metadata_list = try_get( player_response, lambda x: x['multicamera']['playerLegacyMulticameraRenderer']['metadataList'], - compat_str) or try_get( - video_info, lambda x: x['multifeed_metadata_list'][0], compat_str) + compat_str) if multifeed_metadata_list: entries = [] feed_ids = [] @@ -1821,10 +1298,12 @@ class YoutubeIE(YoutubeBaseInfoExtractor): # Unquote should take place before split on comma (,) since textual # fields may contain comma as well (see # https://github.com/ytdl-org/youtube-dl/issues/8536) - feed_data = compat_parse_qs(compat_urllib_parse_unquote_plus(feed)) + feed_data = compat_parse_qs( + compat_urllib_parse_unquote_plus(feed)) def feed_entry(name): - return try_get(feed_data, lambda x: x[name][0], compat_str) + return try_get( + feed_data, lambda x: x[name][0], compat_str) feed_id = feed_entry('id') if not feed_id: @@ -1837,7 +1316,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor): '_type': 'url_transparent', 'ie_key': 'Youtube', 'url': smuggle_url( - '%s://www.youtube.com/watch?v=%s' % (proto, feed_data['id'][0]), + base_url + 'watch?v=' + feed_data['id'][0], {'force_singlefeed': True}), 'title': title, }) @@ -1845,631 +1324,393 @@ class YoutubeIE(YoutubeBaseInfoExtractor): self.to_screen( 'Downloading multifeed video (%s) - add --no-playlist to just download video %s' % (', '.join(feed_ids), video_id)) - return self.playlist_result(entries, video_id, video_title, video_description) + return self.playlist_result( + entries, video_id, video_title, video_description) else: self.to_screen('Downloading just video %s because of --no-playlist' % video_id) - if view_count is None: - view_count = extract_view_count(video_info) - if view_count is None and video_details: - view_count = int_or_none(video_details.get('viewCount')) - if view_count is None and microformat: - view_count = int_or_none(microformat.get('viewCount')) - - if is_live is None: - is_live = bool_or_none(video_details.get('isLive')) - - # Check for "rental" videos - if 'ypc_video_rental_bar_text' in video_info and 'author' not in video_info: - raise ExtractorError('"rental" videos not supported. See https://github.com/ytdl-org/youtube-dl/issues/359 for more information.', expected=True) - - def _extract_filesize(media_url): - return int_or_none(self._search_regex( - r'\bclen[=/](\d+)', media_url, 'filesize', default=None)) - - streaming_formats = try_get(player_response, lambda x: x['streamingData']['formats'], list) or [] - streaming_formats.extend(try_get(player_response, lambda x: x['streamingData']['adaptiveFormats'], list) or []) - - if 'conn' in video_info and video_info['conn'][0].startswith('rtmp'): - self.report_rtmp_download() - formats = [{ - 'format_id': '_rtmp', - 'protocol': 'rtmp', - 'url': video_info['conn'][0], - 'player_url': player_url, - }] - elif not is_live and (streaming_formats or 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/ytdl-org/youtube-dl/issues/343 for more information.', expected=True) - formats = [] - formats_spec = {} - fmt_list = video_info.get('fmt_list', [''])[0] - if fmt_list: - for fmt in fmt_list.split(','): - spec = fmt.split('/') - if len(spec) > 1: - width_height = spec[1].split('x') - if len(width_height) == 2: - formats_spec[spec[0]] = { - 'resolution': spec[1], - 'width': int_or_none(width_height[0]), - 'height': int_or_none(width_height[1]), - } - for fmt in streaming_formats: - itag = str_or_none(fmt.get('itag')) - if not itag: - continue - quality = fmt.get('quality') - quality_label = fmt.get('qualityLabel') or quality - formats_spec[itag] = { - 'asr': int_or_none(fmt.get('audioSampleRate')), - 'filesize': int_or_none(fmt.get('contentLength')), - 'format_note': quality_label, - 'fps': int_or_none(fmt.get('fps')), - 'height': int_or_none(fmt.get('height')), - # bitrate for itag 43 is always 2147483647 - 'tbr': float_or_none(fmt.get('averageBitrate') or fmt.get('bitrate'), 1000) if itag != '43' else None, - 'width': int_or_none(fmt.get('width')), - } + formats = [] + itags = [] + player_url = None + q = qualities(['tiny', 'small', 'medium', 'large', 'hd720', 'hd1080', 'hd1440', 'hd2160', 'hd2880', 'highres']) + streaming_data = player_response.get('streamingData') or {} + streaming_formats = streaming_data.get('formats') or [] + streaming_formats.extend(streaming_data.get('adaptiveFormats') or []) + for fmt in streaming_formats: + if fmt.get('targetDurationSec') or fmt.get('drmFamilies'): + continue - for fmt in streaming_formats: - if fmt.get('drmFamilies') or fmt.get('drm_families'): + fmt_url = fmt.get('url') + if not fmt_url: + sc = compat_parse_qs(fmt.get('signatureCipher')) + fmt_url = url_or_none(try_get(sc, lambda x: x['url'][0])) + encrypted_sig = try_get(sc, lambda x: x['s'][0]) + if not (sc and fmt_url and encrypted_sig): continue - url = url_or_none(fmt.get('url')) - - if not url: - cipher = fmt.get('cipher') or fmt.get('signatureCipher') - if not cipher: + if not player_url: + if not webpage: continue - url_data = compat_parse_qs(cipher) - url = url_or_none(try_get(url_data, lambda x: x['url'][0], compat_str)) - if not url: + player_url = self._search_regex( + r'"(?:PLAYER_JS_URL|jsUrl)"\s*:\s*"([^"]+)"', + webpage, 'player URL', fatal=False) + if not player_url: + continue + signature = self._decrypt_signature(sc['s'][0], video_id, player_url) + sp = try_get(sc, lambda x: x['sp'][0]) or 'signature' + fmt_url += '&' + sp + '=' + signature + + itag = str_or_none(fmt.get('itag')) + if itag: + itags.append(itag) + quality = fmt.get('quality') + dct = { + 'asr': int_or_none(fmt.get('audioSampleRate')), + 'filesize': int_or_none(fmt.get('contentLength')), + 'format_id': itag, + 'format_note': fmt.get('qualityLabel') or quality, + 'fps': int_or_none(fmt.get('fps')), + 'height': int_or_none(fmt.get('height')), + 'quality': q(quality), + 'tbr': float_or_none(fmt.get( + 'averageBitrate') or fmt.get('bitrate'), 1000), + 'url': fmt_url, + 'width': fmt.get('width'), + } + mimetype = fmt.get('mimeType') + if mimetype: + mobj = re.match( + r'((?:[^/]+)/(?:[^;]+))(?:;\s*codecs="([^"]+)")?', mimetype) + if mobj: + dct['ext'] = mimetype2ext(mobj.group(1)) + dct.update(parse_codecs(mobj.group(2))) + if dct.get('acodec') == 'none' or dct.get('vcodec') == 'none': + dct['downloader_options'] = { + # Youtube throttles chunks >~10M + 'http_chunk_size': 10485760, + } + formats.append(dct) + + hls_manifest_url = streaming_data.get('hlsManifestUrl') + if hls_manifest_url: + for f in self._extract_m3u8_formats( + hls_manifest_url, video_id, 'mp4', fatal=False): + itag = self._search_regex( + r'/itag/(\d+)', f['url'], 'itag', default=None) + if itag: + f['format_id'] = itag + formats.append(f) + + if self._downloader.params.get('youtube_include_dash_manifest'): + dash_manifest_url = streaming_data.get('dashManifestUrl') + if dash_manifest_url: + for f in self._extract_mpd_formats( + dash_manifest_url, video_id, fatal=False): + if f['format_id'] in itags: continue - else: - cipher = None - url_data = compat_parse_qs(compat_urllib_parse_urlparse(url).query) + filesize = int_or_none(self._search_regex( + r'/clen/(\d+)', f.get('fragment_base_url') + or f['url'], 'file size', default=None)) + if filesize: + f['filesize'] = filesize + formats.append(f) - stream_type = int_or_none(try_get(url_data, lambda x: x['stream_type'][0])) - # Unsupported FORMAT_STREAM_TYPE_OTF - if stream_type == 3: - continue + if not formats: + if streaming_data.get('licenseInfos'): + raise ExtractorError( + 'This video is DRM protected.', expected=True) + pemr = try_get( + playability_status, + lambda x: x['errorScreen']['playerErrorMessageRenderer'], + dict) or {} + reason = get_text(pemr.get('reason')) or playability_status.get('reason') + subreason = pemr.get('subreason') + if subreason: + subreason = clean_html(get_text(subreason)) + if subreason == 'The uploader has not made this video available in your country.': + countries = microformat.get('availableCountries') + if not countries: + regions_allowed = search_meta('regionsAllowed') + countries = regions_allowed.split(',') if regions_allowed else None + self.raise_geo_restricted( + subreason, countries) + reason += '\n' + subreason + if reason: + raise ExtractorError(reason, expected=True) - format_id = fmt.get('itag') or url_data['itag'][0] - if not format_id: - continue - format_id = compat_str(format_id) - - if cipher: - if 's' in url_data or self._downloader.params.get('youtube_include_dash_manifest', True): - ASSETS_RE = ( - r'<script[^>]+\bsrc=("[^"]+")[^>]+\bname=["\']player_ias/base', - r'"jsUrl"\s*:\s*("[^"]+")', - r'"assets":.+?"js":\s*("[^"]+")') - jsplayer_url_json = self._search_regex( - ASSETS_RE, - embed_webpage if age_gate else video_webpage, - 'JS player URL (1)', default=None) - if not jsplayer_url_json and not age_gate: - # We need the embed website after all - if embed_webpage is None: - embed_url = proto + '://www.youtube.com/embed/%s' % video_id - embed_webpage = self._download_webpage( - embed_url, video_id, 'Downloading embed webpage') - jsplayer_url_json = self._search_regex( - ASSETS_RE, embed_webpage, 'JS player URL') - - player_url = json.loads(jsplayer_url_json) - if player_url is None: - player_url_json = self._search_regex( - r'ytplayer\.config.*?"url"\s*:\s*("[^"]+")', - 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_desc = 'unknown' - else: - player_type, player_version = self._extract_player_info(player_url) - player_desc = '%s player %s' % ('flash' if player_type == 'swf' else 'html5', player_version) - parts_sizes = self._signature_cache_id(encrypted_sig) - self.to_screen('{%s} signature length %s, %s' % - (format_id, parts_sizes, player_desc)) - - signature = self._decrypt_signature( - encrypted_sig, video_id, player_url, age_gate) - sp = try_get(url_data, lambda x: x['sp'][0], compat_str) or 'signature' - url += '&%s=%s' % (sp, signature) - if 'ratebypass' not in url: - url += '&ratebypass=yes' - - dct = { - 'format_id': format_id, - 'url': url, - 'player_url': player_url, - } - if format_id in self._formats: - dct.update(self._formats[format_id]) - if format_id in formats_spec: - dct.update(formats_spec[format_id]) - - # Some itags are not included in DASH manifest thus corresponding formats will - # lack metadata (see https://github.com/ytdl-org/youtube-dl/pull/5993). - # Trying to extract metadata from url_encoded_fmt_stream_map entry. - 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) - - if width is None: - width = int_or_none(fmt.get('width')) - if height is None: - height = int_or_none(fmt.get('height')) - - filesize = int_or_none(url_data.get( - 'clen', [None])[0]) or _extract_filesize(url) - - quality = url_data.get('quality', [None])[0] or fmt.get('quality') - quality_label = url_data.get('quality_label', [None])[0] or fmt.get('qualityLabel') - - tbr = (float_or_none(url_data.get('bitrate', [None])[0], 1000) - or float_or_none(fmt.get('bitrate'), 1000)) if format_id != '43' else None - fps = int_or_none(url_data.get('fps', [None])[0]) or int_or_none(fmt.get('fps')) - - more_fields = { - 'filesize': filesize, - 'tbr': tbr, - 'width': width, - 'height': height, - 'fps': fps, - 'format_note': quality_label or quality, - } - for key, value in more_fields.items(): - if value: - dct[key] = value - type_ = url_data.get('type', [None])[0] or fmt.get('mimeType') - if type_: - type_split = type_.split(';') - kind_ext = type_split[0].split('/') - if len(kind_ext) == 2: - kind, _ = kind_ext - dct['ext'] = mimetype2ext(type_split[0]) - if kind in ('audio', 'video'): - codecs = None - for mobj in re.finditer( - r'(?P<key>[a-zA-Z_-]+)=(?P<quote>["\']?)(?P<val>.+?)(?P=quote)(?:;|$)', type_): - if mobj.group('key') == 'codecs': - codecs = mobj.group('val') - 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) - else: - manifest_url = ( - url_or_none(try_get( - player_response, - lambda x: x['streamingData']['hlsManifestUrl'], - compat_str)) - or url_or_none(try_get( - video_info, lambda x: x['hlsvp'][0], compat_str))) - if manifest_url: - formats = [] - m3u8_formats = self._extract_m3u8_formats( - manifest_url, video_id, 'mp4', fatal=False) - for a_format in m3u8_formats: - itag = self._search_regex( - r'/itag/(\d+)/', a_format['url'], 'itag', default=None) - if itag: - a_format['format_id'] = itag - if itag in self._formats: - dct = self._formats[itag].copy() - dct.update(a_format) - a_format = dct - a_format['player_url'] = player_url - # Accept-Encoding header causes failures in live streams on Youtube and Youtube Gaming - a_format.setdefault('http_headers', {})['Youtubedl-no-compression'] = 'True' - formats.append(a_format) - else: - error_message = extract_unavailable_message() - if not error_message: - reason_list = try_get( - player_response, - lambda x: x['playabilityStatus']['errorScreen']['playerErrorMessageRenderer']['subreason']['runs'], - list) or [] - for reason in reason_list: - if not isinstance(reason, dict): - continue - reason_text = try_get(reason, lambda x: x['text'], compat_str) - if reason_text: - if not error_message: - error_message = '' - error_message += reason_text - if error_message: - error_message = clean_html(error_message) - if not error_message: - error_message = clean_html(try_get( - player_response, lambda x: x['playabilityStatus']['reason'], - compat_str)) - if not error_message: - error_message = clean_html( - try_get(video_info, lambda x: x['reason'][0], compat_str)) - if error_message: - raise ExtractorError(error_message, expected=True) - raise ExtractorError('no conn, hlsvp, hlsManifestUrl or url_encoded_fmt_stream_map information found in video info') - - # uploader - video_uploader = try_get( - video_info, lambda x: x['author'][0], - compat_str) or str_or_none(video_details.get('author')) - 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>[^"]+))">', - video_webpage) - if mobj is not None: - video_uploader_id = mobj.group('uploader_id') - video_uploader_url = mobj.group('uploader_url') - else: - owner_profile_url = url_or_none(microformat.get('ownerProfileUrl')) - if owner_profile_url: - video_uploader_id = self._search_regex( - r'(?:user|channel)/([^/]+)', owner_profile_url, 'uploader id', - default=None) - video_uploader_url = owner_profile_url - - channel_id = ( - str_or_none(video_details.get('channelId')) - or self._html_search_meta( - 'channelId', video_webpage, 'channel id', default=None) - or self._search_regex( - r'data-channel-external-id=(["\'])(?P<id>(?:(?!\1).)+)\1', - video_webpage, 'channel id', default=None, group='id')) - channel_url = 'http://www.youtube.com/channel/%s' % channel_id if channel_id else None + self._sort_formats(formats) - thumbnails = [] - thumbnails_list = try_get( - video_details, lambda x: x['thumbnail']['thumbnails'], list) or [] - for t in thumbnails_list: - if not isinstance(t, dict): - continue - thumbnail_url = url_or_none(t.get('url')) - if not thumbnail_url: - continue - thumbnails.append({ - 'url': thumbnail_url, - 'width': int_or_none(t.get('width')), - 'height': int_or_none(t.get('height')), - }) + keywords = video_details.get('keywords') or [] + if not keywords and webpage: + keywords = [ + unescapeHTML(m.group('content')) + for m in re.finditer(self._meta_regex('og:video:tag'), webpage)] + for keyword in keywords: + if keyword.startswith('yt:stretch='): + w, h = keyword.split('=')[1].split(':') + w, h = int(w), int(h) + if w > 0 and h > 0: + ratio = w / h + for f in formats: + if f.get('vcodec') != 'none': + f['stretched_ratio'] = ratio - if not thumbnails: - video_thumbnail = None - # We try first to get a high quality image: - m_thumb = re.search(r'<span itemprop="thumbnail".*?href="(.*?)">', - video_webpage, re.DOTALL) - if m_thumb is not None: - video_thumbnail = m_thumb.group(1) - thumbnail_url = try_get(video_info, lambda x: x['thumbnail_url'][0], compat_str) - if thumbnail_url: - video_thumbnail = compat_urllib_parse_unquote_plus(thumbnail_url) - if video_thumbnail: - thumbnails.append({'url': video_thumbnail}) - - # upload date - upload_date = self._html_search_meta( - 'datePublished', video_webpage, 'upload date', default=None) - if not upload_date: - upload_date = self._search_regex( - [r'(?s)id="eow-date.*?>(.*?)</span>', - r'(?:id="watch-uploader-info".*?>.*?|["\']simpleText["\']\s*:\s*["\'])(?:Published|Uploaded|Streamed live|Started) on (.+?)[<"\']'], - video_webpage, 'upload date', default=None) - if not upload_date: - upload_date = microformat.get('publishDate') or microformat.get('uploadDate') - upload_date = unified_strdate(upload_date) - - video_license = self._html_search_regex( - r'<h4[^>]+class="title"[^>]*>\s*License\s*</h4>\s*<ul[^>]*>\s*<li>(.+?)</li', - video_webpage, 'license', default=None) - - m_music = re.search( - 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_creator = clean_html(m_music.group('creator')) + thumbnails = [] + for container in (video_details, microformat): + for thumbnail in (try_get( + container, + lambda x: x['thumbnail']['thumbnails'], list) or []): + thumbnail_url = thumbnail.get('url') + if not thumbnail_url: + continue + thumbnails.append({ + 'height': int_or_none(thumbnail.get('height')), + 'url': thumbnail_url, + 'width': int_or_none(thumbnail.get('width')), + }) + if thumbnails: + break else: - video_alt_title = video_creator = None + thumbnail = search_meta(['og:image', 'twitter:image']) + if thumbnail: + thumbnails = [{'url': thumbnail}] + + category = microformat.get('category') or search_meta('genre') + channel_id = video_details.get('channelId') \ + or microformat.get('externalChannelId') \ + or search_meta('channelId') + duration = int_or_none( + video_details.get('lengthSeconds') + or microformat.get('lengthSeconds')) \ + or parse_duration(search_meta('duration')) + is_live = video_details.get('isLive') + owner_profile_url = microformat.get('ownerProfileUrl') + + info = { + 'id': video_id, + 'title': self._live_title(video_title) if is_live else video_title, + 'formats': formats, + 'thumbnails': thumbnails, + 'description': video_description, + 'upload_date': unified_strdate( + microformat.get('uploadDate') + or search_meta('uploadDate')), + 'uploader': video_details['author'], + 'uploader_id': self._search_regex(r'/(?:channel|user)/([^/?&#]+)', owner_profile_url, 'uploader id') if owner_profile_url else None, + 'uploader_url': owner_profile_url, + 'channel_id': channel_id, + 'channel_url': 'https://www.youtube.com/channel/' + channel_id if channel_id else None, + 'duration': duration, + 'view_count': int_or_none( + video_details.get('viewCount') + or microformat.get('viewCount') + or search_meta('interactionCount')), + 'average_rating': float_or_none(video_details.get('averageRating')), + 'age_limit': 18 if ( + microformat.get('isFamilySafe') is False + or search_meta('isFamilyFriendly') == 'false' + or search_meta('og:restrictions:age') == '18+') else 0, + 'webpage_url': webpage_url, + 'categories': [category] if category else None, + 'tags': keywords, + 'is_live': is_live, + } - def extract_meta(field): - return self._html_search_regex( - r'<h4[^>]+class="title"[^>]*>\s*%s\s*</h4>\s*<ul[^>]*>\s*<li>(.+?)</li>\s*' % field, - video_webpage, field, default=None) + pctr = try_get( + player_response, + lambda x: x['captions']['playerCaptionsTracklistRenderer'], dict) + if pctr: + def process_language(container, base_url, caption, query): + lang_subs = [] + for fmt in self._SUBTITLE_FORMATS: + query.update({ + 'fmt': fmt, + }) + lang_subs.append({ + 'ext': fmt, + 'url': update_url_query(base_url, query), + }) + subtitles[caption['languageCode']] = lang_subs + + subtitles = {} + for caption_track in pctr['captionTracks']: + base_url = caption_track['baseUrl'] + if caption_track.get('kind') != 'asr': + lang_subs = [] + for fmt in self._SUBTITLE_FORMATS: + lang_subs.append({ + 'ext': fmt, + 'url': update_url_query(base_url, { + 'fmt': fmt, + }), + }) + subtitles[caption_track['languageCode']] = lang_subs + continue + automatic_captions = {} + for translation_language in pctr['translationLanguages']: + translation_language_code = translation_language['languageCode'] + lang_subs = [] + for fmt in self._SUBTITLE_FORMATS: + lang_subs.append({ + 'ext': fmt, + 'url': update_url_query(base_url, { + 'fmt': fmt, + 'tlang': translation_language_code, + }), + }) + automatic_captions[translation_language_code] = lang_subs + info['automatic_captions'] = automatic_captions + info['subtitles'] = subtitles - track = extract_meta('Song') - artist = extract_meta('Artist') - album = extract_meta('Album') + parsed_url = compat_urllib_parse_urlparse(url) + for component in [parsed_url.fragment, parsed_url.query]: + query = compat_parse_qs(component) + for k, v in query.items(): + for d_k, s_ks in [('start', ('start', 't')), ('end', ('end',))]: + d_k += '_time' + if d_k not in info and k in s_ks: + info[d_k] = parse_duration(query[k][0]) - # Youtube Music Auto-generated description - release_date = release_year = None if video_description: mobj = re.search(r'(?s)(?P<track>[^·\n]+)·(?P<artist>[^\n]+)\n+(?P<album>[^\n]+)(?:.+?℗\s*(?P<release_year>\d{4})(?!\d))?(?:.+?Released on\s*:\s*(?P<release_date>\d{4}-\d{2}-\d{2}))?(.+?\nArtist\s*:\s*(?P<clean_artist>[^\n]+))?.+\nAuto-generated by YouTube\.\s*$', video_description) if mobj: - if not track: - track = mobj.group('track').strip() - if not artist: - artist = mobj.group('clean_artist') or ', '.join(a.strip() for a in mobj.group('artist').split('·')) - if not album: - album = mobj.group('album'.strip()) release_year = mobj.group('release_year') release_date = mobj.group('release_date') if release_date: release_date = release_date.replace('-', '') if not release_year: - release_year = int(release_date[:4]) - if release_year: - release_year = int(release_year) + release_year = release_date[:4] + info.update({ + 'album': mobj.group('album'.strip()), + 'artist': mobj.group('clean_artist') or ', '.join(a.strip() for a in mobj.group('artist').split('·')), + 'track': mobj.group('track').strip(), + 'release_date': release_date, + 'release_year': int(release_year), + }) - yt_initial_data = self._extract_yt_initial_data(video_id, video_webpage) - contents = try_get(yt_initial_data, lambda x: x['contents']['twoColumnWatchNextResults']['results']['results']['contents'], list) or [] - for content in contents: - rows = try_get(content, lambda x: x['videoSecondaryInfoRenderer']['metadataRowContainer']['metadataRowContainerRenderer']['rows'], list) or [] - multiple_songs = False - for row in rows: - if try_get(row, lambda x: x['metadataRowRenderer']['hasDividerLine']) is True: - multiple_songs = True - break - for row in rows: - mrr = row.get('metadataRowRenderer') or {} - mrr_title = try_get( - mrr, lambda x: x['title']['simpleText'], compat_str) - mrr_contents = try_get( - mrr, lambda x: x['contents'][0], dict) or {} - mrr_contents_text = try_get(mrr_contents, [lambda x: x['simpleText'], lambda x: x['runs'][0]['text']], compat_str) - if not (mrr_title and mrr_contents_text): + initial_data = None + if webpage: + initial_data = self._extract_yt_initial_variable( + webpage, self._YT_INITIAL_DATA_RE, video_id, + 'yt initial data') + if not initial_data: + initial_data = self._call_api( + 'next', {'videoId': video_id}, video_id, fatal=False) + + if initial_data: + for engagment_pannel in (initial_data.get('engagementPanels') or []): + contents = try_get( + engagment_pannel, lambda x: x['engagementPanelSectionListRenderer']['content']['macroMarkersListRenderer']['contents'], + list) + if not contents: continue - if mrr_title == 'License': - video_license = mrr_contents_text - elif not multiple_songs: - if mrr_title == 'Album': - album = mrr_contents_text - elif mrr_title == 'Artist': - artist = mrr_contents_text - elif mrr_title == 'Song': - track = mrr_contents_text - - m_episode = re.search( - r'<div[^>]+id="watch7-headline"[^>]*>\s*<span[^>]*>.*?>(?P<series>[^<]+)</a></b>\s*S(?P<season>\d+)\s*•\s*E(?P<episode>\d+)</span>', - video_webpage) - if m_episode: - series = unescapeHTML(m_episode.group('series')) - season_number = int(m_episode.group('season')) - episode_number = int(m_episode.group('episode')) - else: - series = season_number = episode_number = None - - m_cat_container = self._search_regex( - r'(?s)<h4[^>]*>\s*Category\s*</h4>\s*<ul[^>]*>(.*?)</ul>', - video_webpage, 'categories', default=None) - category = None - if m_cat_container: - category = self._html_search_regex( - r'(?s)<a[^<]+>(.*?)</a>', m_cat_container, 'category', - default=None) - if not category: - category = try_get( - microformat, lambda x: x['category'], compat_str) - video_categories = None if category is None else [category] - - video_tags = [ - unescapeHTML(m.group('content')) - for m in re.finditer(self._meta_regex('og:video:tag'), video_webpage)] - if not video_tags: - video_tags = try_get(video_details, lambda x: x['keywords'], list) - - def _extract_count(count_name): - return str_to_int(self._search_regex( - (r'-%s-button[^>]+><span[^>]+class="yt-uix-button-content"[^>]*>([\d,]+)</span>' % re.escape(count_name), - r'["\']label["\']\s*:\s*["\']([\d,.]+)\s+%ss["\']' % re.escape(count_name)), - video_webpage, count_name, default=None)) - - like_count = _extract_count('like') - dislike_count = _extract_count('dislike') - - if view_count is None: - view_count = str_to_int(self._search_regex( - r'<[^>]+class=["\']watch-view-count[^>]+>\s*([\d,\s]+)', video_webpage, - 'view count', default=None)) - - average_rating = ( - float_or_none(video_details.get('averageRating')) - or try_get(video_info, lambda x: float_or_none(x['avg_rating'][0]))) - - # subtitles - video_subtitles = self.extract_subtitles(video_id, video_webpage) - automatic_captions = self.extract_automatic_captions(video_id, player_response, ytplayer_config) - - video_duration = try_get( - video_info, lambda x: int_or_none(x['length_seconds'][0])) - if not video_duration: - video_duration = int_or_none(video_details.get('lengthSeconds')) - if not video_duration: - video_duration = parse_duration(self._html_search_meta( - 'duration', video_webpage, 'video duration')) - - # annotations - video_annotations = None - if self._downloader.params.get('writeannotations', False): - xsrf_token = None - ytcfg = self._extract_ytcfg(video_id, video_webpage) - if ytcfg: - xsrf_token = try_get(ytcfg, lambda x: x['XSRF_TOKEN'], compat_str) - if not xsrf_token: - xsrf_token = self._search_regex( - r'([\'"])XSRF_TOKEN\1\s*:\s*([\'"])(?P<xsrf_token>(?:(?!\2).)+)\2', - video_webpage, 'xsrf token', group='xsrf_token', fatal=False) - invideo_url = try_get( - player_response, lambda x: x['annotations'][0]['playerAnnotationsUrlsRenderer']['invideoUrl'], compat_str) - if xsrf_token and invideo_url: - xsrf_field_name = None - if ytcfg: - xsrf_field_name = try_get(ytcfg, lambda x: x['XSRF_FIELD_NAME'], compat_str) - if not xsrf_field_name: - xsrf_field_name = self._search_regex( - r'([\'"])XSRF_FIELD_NAME\1\s*:\s*([\'"])(?P<xsrf_field_name>\w+)\2', - video_webpage, 'xsrf field name', - group='xsrf_field_name', default='session_token') - video_annotations = self._download_webpage( - self._proto_relative_url(invideo_url), - video_id, note='Downloading annotations', - errnote='Unable to download video annotations', fatal=False, - data=urlencode_postdata({xsrf_field_name: xsrf_token})) - - chapters = self._extract_chapters(video_webpage, description_original, video_id, video_duration) - - # Look for the DASH manifest - if self._downloader.params.get('youtube_include_dash_manifest', True): - dash_mpd_fatal = True - for mpd_url in dash_mpds: - dash_formats = {} - try: - def decrypt_sig(mobj): - s = mobj.group(1) - dec_s = self._decrypt_signature(s, video_id, player_url, age_gate) - return '/signature/%s' % dec_s - - mpd_url = re.sub(r'/s/([a-fA-F0-9\.]+)', decrypt_sig, mpd_url) - - 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.get('fragment_base_url') or 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 - # Additional DASH manifests may end up in HTTP Error 403 therefore - # allow them to fail without bug report message if we already have - # some DASH manifest succeeded. This is temporary workaround to reduce - # burst of bug reports until we figure out the reason and whether it - # can be fixed at all. - dash_mpd_fatal = False - except (ExtractorError, KeyError) as e: - self.report_warning( - 'Skipping DASH manifest: %r' % e, video_id) - if dash_formats: - # Remove the formats we found through non-DASH, they - # contain less info and it can be wrong, because we use - # fixed values (for example the resolution). See - # https://github.com/ytdl-org/youtube-dl/issues/5774 for an - # example. - formats = [f for f in formats if f['format_id'] not in dash_formats.keys()] - formats.extend(dash_formats.values()) - - # Check for malformed aspect ratio - stretched_m = re.search( - r'<meta\s+property="og:video:tag".*?content="yt:stretch=(?P<w>[0-9]+):(?P<h>[0-9]+)">', - video_webpage) - if stretched_m: - w = float(stretched_m.group('w')) - h = float(stretched_m.group('h')) - # yt:stretch may hold invalid ratio data (e.g. for Q39EVAstoRM ratio is 17:0). - # We will only process correct ratios. - if w > 0 and h > 0: - ratio = w / h - for f in formats: - if f.get('vcodec') != 'none': - f['stretched_ratio'] = ratio - - if not formats: - if 'reason' in video_info: - if 'The uploader has not made this video available in your country.' in video_info['reason']: - regions_allowed = self._html_search_meta( - 'regionsAllowed', video_webpage, default=None) - 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' % reason, - expected=True, video_id=video_id) - if video_info.get('license_info') or try_get(player_response, lambda x: x['streamingData']['licenseInfos']): - raise ExtractorError('This video is DRM protected.', expected=True) - - self._sort_formats(formats) - self.mark_watched(video_id, video_info, player_response) - - return { - 'id': video_id, - 'uploader': video_uploader, - 'uploader_id': video_uploader_id, - 'uploader_url': video_uploader_url, - 'channel_id': channel_id, - 'channel_url': channel_url, - 'upload_date': upload_date, - 'license': video_license, - 'creator': video_creator or artist, - 'title': video_title, - 'alt_title': video_alt_title or track, - 'thumbnails': thumbnails, - 'description': video_description, - 'categories': video_categories, - 'tags': video_tags, - 'subtitles': video_subtitles, - 'automatic_captions': automatic_captions, - 'duration': video_duration, - 'age_limit': 18 if age_gate else 0, - 'annotations': video_annotations, - 'chapters': chapters, - 'webpage_url': proto + '://www.youtube.com/watch?v=%s' % video_id, - 'view_count': view_count, - 'like_count': like_count, - 'dislike_count': dislike_count, - 'average_rating': average_rating, - 'formats': formats, - 'is_live': is_live, - 'start_time': start_time, - 'end_time': end_time, - 'series': series, - 'season_number': season_number, - 'episode_number': episode_number, - 'track': track, - 'artist': artist, - 'album': album, - 'release_date': release_date, - 'release_year': release_year, - } + def chapter_time(mmlir): + return parse_duration(mmlir.get( + get_text(mmlir.get('timeDescription')))) + + chapters = [] + for next_num, content in enumerate(contents, start=1): + mmlir = content.get('macroMarkersListItemRenderer') or {} + start_time = chapter_time(mmlir) + end_time = chapter_time(try_get( + contents, lambda x: x[next_num]['macroMarkersListItemRenderer'])) \ + if next_num < len(contents) else duration + if not (start_time and end_time): + continue + chapters.append({ + 'start_time': start_time, + 'end_time': end_time, + 'title': get_text(mmlir.get('title')), + }) + info['chapters'] = chapters + + contents = try_get( + initial_data, + lambda x: x['contents']['twoColumnWatchNextResults']['results']['results']['contents'], + list) or [] + for content in contents: + vpir = content.get('videoPrimaryInfoRenderer') + if vpir: + stl = vpir.get('superTitleLink') + if stl: + stl = get_text(stl) + if try_get( + vpir, + lambda x: x['superTitleIcon']['iconType']) == 'LOCATION_PIN': + info['location'] = stl + else: + mobj = re.search(r'(.+?)\s*S(\d+)\s*•\s*E(\d+)', stl) + if mobj: + info.update({ + 'series': mobj.group(1), + 'season_number': int(mobj.group(2)), + 'episode_number': int(mobj.group(3)), + }) + for tlb in (try_get( + vpir, + lambda x: x['videoActions']['menuRenderer']['topLevelButtons'], + list) or []): + tbr = tlb.get('toggleButtonRenderer') or {} + for getter, regex in [( + lambda x: x['defaultText']['accessibility']['accessibilityData'], + r'(?P<count>[\d,]+)\s*(?P<type>(?:dis)?like)'), ([ + lambda x: x['accessibility'], + lambda x: x['accessibilityData']['accessibilityData'], + ], r'(?P<type>(?:dis)?like) this video along with (?P<count>[\d,]+) other people')]: + label = (try_get(tbr, getter, dict) or {}).get('label') + if label: + mobj = re.match(regex, label) + if mobj: + info[mobj.group('type') + '_count'] = str_to_int(mobj.group('count')) + break + sbr_tooltip = try_get( + vpir, lambda x: x['sentimentBar']['sentimentBarRenderer']['tooltip']) + if sbr_tooltip: + like_count, dislike_count = sbr_tooltip.split(' / ') + info.update({ + 'like_count': str_to_int(like_count), + 'dislike_count': str_to_int(dislike_count), + }) + vsir = content.get('videoSecondaryInfoRenderer') + if vsir: + info['channel'] = get_text(try_get( + vsir, + lambda x: x['owner']['videoOwnerRenderer']['title'], + compat_str)) + rows = try_get( + vsir, + lambda x: x['metadataRowContainer']['metadataRowContainerRenderer']['rows'], + list) or [] + multiple_songs = False + for row in rows: + if try_get(row, lambda x: x['metadataRowRenderer']['hasDividerLine']) is True: + multiple_songs = True + break + for row in rows: + mrr = row.get('metadataRowRenderer') or {} + mrr_title = mrr.get('title') + if not mrr_title: + continue + mrr_title = get_text(mrr['title']) + mrr_contents_text = get_text(mrr['contents'][0]) + if mrr_title == 'License': + info['license'] = mrr_contents_text + elif not multiple_songs: + if mrr_title == 'Album': + info['album'] = mrr_contents_text + elif mrr_title == 'Artist': + info['artist'] = mrr_contents_text + elif mrr_title == 'Song': + info['track'] = mrr_contents_text + + for s_k, d_k in [('artist', 'creator'), ('track', 'alt_title')]: + v = info.get(s_k) + if v: + info[d_k] = v + + self.mark_watched(video_id, player_response) + + return info class YoutubeTabIE(YoutubeBaseInfoExtractor): |