# coding: utf-8
+import collections
+import errno
+import io
+import itertools
import json
-import netrc
+import os.path
import re
-import socket
-import itertools
+import struct
+import traceback
+import zlib
from .common import InfoExtractor, SearchInfoExtractor
+from .subtitles import SubtitlesInfoExtractor
+from ..jsinterp import JSInterpreter
from ..utils import (
- compat_http_client,
+ compat_chr,
compat_parse_qs,
- compat_urllib_error,
compat_urllib_parse,
compat_urllib_request,
+ compat_urlparse,
compat_str,
clean_html,
+ get_cachedir,
get_element_by_id,
+ get_element_by_attribute,
ExtractorError,
+ int_or_none,
+ PagedList,
unescapeHTML,
unified_strdate,
orderedSet,
+ write_json_file,
+ uppercase_escape,
)
class YoutubeBaseInfoExtractor(InfoExtractor):
"""Provide base functions for Youtube extractors"""
_LOGIN_URL = 'https://accounts.google.com/ServiceLogin'
_LANG_URL = r'https://www.youtube.com/?hl=en&persist_hl=1&gl=US&persist_gl=1&opt_out_ackd=1'
- _AGE_URL = 'http://www.youtube.com/verify_age?next_url=/&gl=US&hl=en'
+ _AGE_URL = 'https://www.youtube.com/verify_age?next_url=/&gl=US&hl=en'
_NETRC_MACHINE = 'youtube'
# If True it will raise an error if no login info is provided
_LOGIN_REQUIRED = False
- def report_lang(self):
- """Report attempt to set language."""
- self.to_screen(u'Setting language')
-
def _set_language(self):
- request = compat_urllib_request.Request(self._LANG_URL)
- try:
- self.report_lang()
- compat_urllib_request.urlopen(request).read()
- except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
- self._downloader.report_warning(u'unable to set language: %s' % compat_str(err))
- return False
- return True
+ return bool(self._download_webpage(
+ self._LANG_URL, None,
+ note=u'Setting language', errnote='unable to set language',
+ fatal=False))
def _login(self):
(username, password) = self._get_login_info()
raise ExtractorError(u'No login info available, needed for using %s.' % self.IE_NAME, expected=True)
return False
- request = compat_urllib_request.Request(self._LOGIN_URL)
- try:
- login_page = compat_urllib_request.urlopen(request).read().decode('utf-8')
- except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
- self._downloader.report_warning(u'unable to fetch login page: %s' % compat_str(err))
- return False
+ login_page = self._download_webpage(
+ self._LOGIN_URL, None,
+ note=u'Downloading login page',
+ errnote=u'unable to fetch login page', fatal=False)
+ if login_page is False:
+ return
- galx = None
- dsh = None
- match = re.search(re.compile(r'<input.+?name="GALX".+?value="(.+?)"', re.DOTALL), login_page)
- if match:
- galx = match.group(1)
- match = re.search(re.compile(r'<input.+?name="dsh".+?value="(.+?)"', re.DOTALL), login_page)
- if match:
- dsh = match.group(1)
+ galx = self._search_regex(r'(?s)<input.+?name="GALX".+?value="(.+?)"',
+ login_page, u'Login GALX parameter')
# Log in
login_form_strs = {
u'checkConnection': u'',
u'checkedDomains': u'youtube',
u'dnConn': u'',
- u'dsh': dsh,
u'pstMsg': u'0',
u'rmShown': u'1',
u'secTok': u'',
# chokes on unicode
login_form = dict((k.encode('utf-8'), v.encode('utf-8')) for k,v in login_form_strs.items())
login_data = compat_urllib_parse.urlencode(login_form).encode('ascii')
- request = compat_urllib_request.Request(self._LOGIN_URL, login_data)
- try:
- self.report_login()
- login_results = compat_urllib_request.urlopen(request).read().decode('utf-8')
- if re.search(r'(?i)<form[^>]* id="gaia_loginform"', login_results) is not None:
- self._downloader.report_warning(u'unable to log in: bad username or password')
- return False
- except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
- self._downloader.report_warning(u'unable to log in: %s' % compat_str(err))
+
+ req = compat_urllib_request.Request(self._LOGIN_URL, login_data)
+ login_results = self._download_webpage(
+ req, None,
+ note=u'Logging in', errnote=u'unable to log in', fatal=False)
+ if login_results is False:
+ return False
+ if re.search(r'(?i)<form[^>]* id="gaia_loginform"', login_results) is not None:
+ self._downloader.report_warning(u'unable to log in: bad username or password')
return False
return True
def _confirm_age(self):
age_form = {
- 'next_url': '/',
- 'action_confirm': 'Confirm',
- }
- request = compat_urllib_request.Request(self._AGE_URL, compat_urllib_parse.urlencode(age_form))
- try:
- self.report_age_confirmation()
- compat_urllib_request.urlopen(request).read().decode('utf-8')
- except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
- raise ExtractorError(u'Unable to confirm age: %s' % compat_str(err))
+ 'next_url': '/',
+ 'action_confirm': 'Confirm',
+ }
+ req = compat_urllib_request.Request(self._AGE_URL,
+ compat_urllib_parse.urlencode(age_form).encode('ascii'))
+
+ self._download_webpage(
+ req, None,
+ note=u'Confirming age', errnote=u'Unable to confirm age')
return True
def _real_initialize(self):
return
self._confirm_age()
-class YoutubeIE(YoutubeBaseInfoExtractor):
+
+class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor):
IE_DESC = u'YouTube.com'
- _VALID_URL = r"""^
+ _VALID_URL = r"""(?x)^
(
- (?:https?://)? # http(s):// (optional)
- (?:youtu\.be/|(?:\w+\.)?youtube(?:-nocookie)?\.com/|
- tube\.majestyc\.net/) # the various hostnames, with wildcard subdomains
+ (?:https?://|//)? # http(s):// or protocol-independent URL (optional)
+ (?:(?:(?:(?:\w+\.)?[yY][oO][uU][tT][uU][bB][eE](?:-nocookie)?\.com/|
+ (?:www\.)?deturl\.com/www\.youtube\.com/|
+ (?:www\.)?pwnyoutube\.com/|
+ (?:www\.)?yourepeat\.com/|
+ tube\.majestyc\.net/|
+ youtube\.googleapis\.com/) # the various hostnames, with wildcard subdomains
(?:.*?\#/)? # handle anchor (#/) redirect urls
(?: # the various things that can precede the ID:
(?:(?:v|embed|e)/) # v/ or embed/ or e/
|(?: # or the v= param in all its forms
- (?:watch|movie(?:_popup)?(?:\.php)?)? # preceding watch(_popup|.php) or nothing (like /?v=xxxx)
+ (?:(?:watch|movie)(?:_popup)?(?:\.php)?/?)? # preceding watch(_popup|.php) or nothing (like /?v=xxxx)
(?:\?|\#!?) # the params delimiter ? or # or #!
(?:.*?&)? # any other preceding param (like /?s=tuff&v=xxxx)
v=
)
- )? # optional -> youtube.com/xxxx is OK
+ ))
+ |youtu\.be/ # just youtu.be/xxxx
+ |https?://(?:www\.)?cleanvideosearch\.com/media/action/yt/watch\?videoId=
+ )
)? # all until now is optional -> you can pass the naked ID
- ([0-9A-Za-z_-]+) # here is it! the YouTube video ID
+ ([0-9A-Za-z_-]{11}) # here is it! the YouTube video ID
(?(1).+)? # if we found the ID, everything can follow
$"""
_NEXT_URL_RE = r'[\?&]next_url=([^&]+)'
- # Listed in order of quality
- _available_formats = ['38', '37', '46', '22', '45', '35', '44', '34', '18', '43', '6', '5', '17', '13',
- '95', '94', '93', '92', '132', '151',
- '85', '84', '102', '83', '101', '82', '100',
- ]
- _available_formats_prefer_free = ['38', '46', '37', '45', '22', '44', '35', '43', '34', '18', '6', '5', '17', '13',
- '95', '94', '93', '92', '132', '151',
- '85', '102', '84', '101', '83', '100', '82',
- ]
- _video_extensions = {
- '13': '3gp',
- '17': 'mp4',
- '18': 'mp4',
- '22': 'mp4',
- '37': 'mp4',
- '38': 'mp4',
- '43': 'webm',
- '44': 'webm',
- '45': 'webm',
- '46': 'webm',
+ _formats = {
+ '5': {'ext': 'flv', 'width': 400, 'height': 240},
+ '6': {'ext': 'flv', 'width': 450, 'height': 270},
+ '13': {'ext': '3gp'},
+ '17': {'ext': '3gp', 'width': 176, 'height': 144},
+ '18': {'ext': 'mp4', 'width': 640, 'height': 360},
+ '22': {'ext': 'mp4', 'width': 1280, 'height': 720},
+ '34': {'ext': 'flv', 'width': 640, 'height': 360},
+ '35': {'ext': 'flv', 'width': 854, 'height': 480},
+ '36': {'ext': '3gp', 'width': 320, 'height': 240},
+ '37': {'ext': 'mp4', 'width': 1920, 'height': 1080},
+ '38': {'ext': 'mp4', 'width': 4096, 'height': 3072},
+ '43': {'ext': 'webm', 'width': 640, 'height': 360},
+ '44': {'ext': 'webm', 'width': 854, 'height': 480},
+ '45': {'ext': 'webm', 'width': 1280, 'height': 720},
+ '46': {'ext': 'webm', 'width': 1920, 'height': 1080},
+
# 3d videos
- '82': 'mp4',
- '83': 'mp4',
- '84': 'mp4',
- '85': 'mp4',
- '100': 'webm',
- '101': 'webm',
- '102': 'webm',
-
- # videos that use m3u8
- '92': 'mp4',
- '93': 'mp4',
- '94': 'mp4',
- '95': 'mp4',
- '96': 'mp4',
- '132': 'mp4',
- '151': 'mp4',
+ '82': {'ext': 'mp4', 'height': 360, 'format_note': '3D', 'preference': -20},
+ '83': {'ext': 'mp4', 'height': 480, 'format_note': '3D', 'preference': -20},
+ '84': {'ext': 'mp4', 'height': 720, 'format_note': '3D', 'preference': -20},
+ '85': {'ext': 'mp4', 'height': 1080, 'format_note': '3D', 'preference': -20},
+ '100': {'ext': 'webm', 'height': 360, 'format_note': '3D', 'preference': -20},
+ '101': {'ext': 'webm', 'height': 480, 'format_note': '3D', 'preference': -20},
+ '102': {'ext': 'webm', 'height': 720, 'format_note': '3D', 'preference': -20},
+
+ # Apple HTTP Live Streaming
+ '92': {'ext': 'mp4', 'height': 240, 'format_note': 'HLS', 'preference': -10},
+ '93': {'ext': 'mp4', 'height': 360, 'format_note': 'HLS', 'preference': -10},
+ '94': {'ext': 'mp4', 'height': 480, 'format_note': 'HLS', 'preference': -10},
+ '95': {'ext': 'mp4', 'height': 720, 'format_note': 'HLS', 'preference': -10},
+ '96': {'ext': 'mp4', 'height': 1080, 'format_note': 'HLS', 'preference': -10},
+ '132': {'ext': 'mp4', 'height': 240, 'format_note': 'HLS', 'preference': -10},
+ '151': {'ext': 'mp4', 'height': 72, 'format_note': 'HLS', 'preference': -10},
+
+ # DASH mp4 video
+ '133': {'ext': 'mp4', 'height': 240, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
+ '134': {'ext': 'mp4', 'height': 360, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
+ '135': {'ext': 'mp4', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
+ '136': {'ext': 'mp4', 'height': 720, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
+ '137': {'ext': 'mp4', 'height': 1080, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
+ '138': {'ext': 'mp4', 'height': 2160, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
+ '160': {'ext': 'mp4', 'height': 144, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
+ '264': {'ext': 'mp4', 'height': 1440, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
+
+ # Dash mp4 audio
+ '139': {'ext': 'm4a', 'format_note': 'DASH audio', 'vcodec': 'none', 'abr': 48, 'preference': -50},
+ '140': {'ext': 'm4a', 'format_note': 'DASH audio', 'vcodec': 'none', 'abr': 128, 'preference': -50},
+ '141': {'ext': 'm4a', 'format_note': 'DASH audio', 'vcodec': 'none', 'abr': 256, 'preference': -50},
+
+ # Dash webm
+ '167': {'ext': 'webm', 'height': 360, 'width': 640, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
+ '168': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
+ '169': {'ext': 'webm', 'height': 720, 'width': 1280, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
+ '170': {'ext': 'webm', 'height': 1080, 'width': 1920, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
+ '218': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
+ '219': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40},
+ '242': {'ext': 'webm', 'height': 240, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
+ '243': {'ext': 'webm', 'height': 360, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
+ '244': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
+ '245': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
+ '246': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
+ '247': {'ext': 'webm', 'height': 720, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
+ '248': {'ext': 'webm', 'height': 1080, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
+ '271': {'ext': 'webm', 'height': 1440, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
+ '272': {'ext': 'webm', 'height': 2160, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40},
+
+ # Dash webm audio
+ '171': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH audio', 'abr': 48, 'preference': -50},
+ '172': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH audio', 'abr': 256, 'preference': -50},
+
+ # RTMP (unnamed)
+ '_rtmp': {'protocol': 'rtmp'},
}
- _video_dimensions = {
- '5': '240x400',
- '6': '???',
- '13': '???',
- '17': '144x176',
- '18': '360x640',
- '22': '720x1280',
- '34': '360x640',
- '35': '480x854',
- '37': '1080x1920',
- '38': '3072x4096',
- '43': '360x640',
- '44': '480x854',
- '45': '720x1280',
- '46': '1080x1920',
- '82': '360p',
- '83': '480p',
- '84': '720p',
- '85': '1080p',
- '92': '240p',
- '93': '360p',
- '94': '480p',
- '95': '720p',
- '96': '1080p',
- '100': '360p',
- '101': '480p',
- '102': '720p',
- '132': '240p',
- '151': '72p',
- }
- _3d_itags = ['85', '84', '102', '83', '101', '82', '100']
+
IE_NAME = u'youtube'
_TESTS = [
{
u"uploader": u"Philipp Hagemeister",
u"uploader_id": u"phihag",
u"upload_date": u"20121002",
- u"description": u"test chars: \"'/\\Ƥāš\n\nThis is a test video for youtube-dl.\n\nFor more information, contact phihag@phihag.de ."
- }
- },
- {
- u"url": u"http://www.youtube.com/watch?v=1ltcDfZMA3U",
- u"file": u"1ltcDfZMA3U.flv",
- u"note": u"Test VEVO video (#897)",
- u"info_dict": {
- u"upload_date": u"20070518",
- u"title": u"Maps - It Will Find You",
- u"description": u"Music video by Maps performing It Will Find You.",
- u"uploader": u"MuteUSA",
- u"uploader_id": u"MuteUSA"
+ u"description": u"test chars: \"'/\\Ƥāš\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 .",
+ u"categories": [u'Science & Technology'],
}
},
{
u"info_dict": {
u"upload_date": u"20120506",
u"title": u"Icona Pop - I Love It (feat. Charli XCX) [OFFICIAL VIDEO]",
- u"description": u"md5:b085c9804f5ab69f4adea963a2dceb3c",
- u"uploader": u"IconaPop",
+ u"description": u"md5:fea86fda2d5a5784273df5c7cc994d9f",
+ u"uploader": u"Icona Pop",
u"uploader_id": u"IconaPop"
}
},
}
},
{
- u'url': u'https://www.youtube.com/watch?v=TGi3HqYrWHE',
- u'file': u'TGi3HqYrWHE.mp4',
- u'note': u'm3u8 video',
+ u"url": u"//www.YouTube.com/watch?v=yZIXLfi8CZQ",
+ u"file": u"yZIXLfi8CZQ.mp4",
+ u"note": u"Embed-only video (#1746)",
+ u"info_dict": {
+ u"upload_date": u"20120608",
+ u"title": u"Principal Sexually Assaults A Teacher - Episode 117 - 8th June 2012",
+ u"description": u"md5:09b78bd971f1e3e289601dfba15ca4f7",
+ u"uploader": u"SET India",
+ u"uploader_id": u"setindia"
+ }
+ },
+ {
+ u"url": u"http://www.youtube.com/watch?v=a9LDPn-MO4I",
+ u"file": u"a9LDPn-MO4I.m4a",
+ u"note": u"256k DASH audio (format 141) via DASH manifest",
+ u"info_dict": {
+ u"upload_date": "20121002",
+ u"uploader_id": "8KVIDEO",
+ u"description": "No description available.",
+ u"uploader": "8KVIDEO",
+ u"title": "UHDTV TEST 8K VIDEO.mp4"
+ },
+ u"params": {
+ u"youtube_include_dash_manifest": True,
+ u"format": "141",
+ },
+ },
+ # DASH manifest with encrypted signature
+ {
+ u'url': u'https://www.youtube.com/watch?v=IB3lcPjvWLA',
u'info_dict': {
- u'title': u'Triathlon - Men - London 2012 Olympic Games',
- u'description': u'- Men - TR02 - Triathlon - 07 August 2012 - London 2012 Olympic Games',
- u'uploader': u'olympic',
- u'upload_date': u'20120807',
- u'uploader_id': u'olympic',
+ u'id': u'IB3lcPjvWLA',
+ u'ext': u'm4a',
+ u'title': u'Afrojack - The Spark ft. Spree Wilson',
+ u'description': u'md5:9717375db5a9a3992be4668bbf3bc0a8',
+ u'uploader': u'AfrojackVEVO',
+ u'uploader_id': u'AfrojackVEVO',
+ u'upload_date': u'20131011',
},
- u'params': {
- u'skip_download': True,
+ u"params": {
+ u'youtube_include_dash_manifest': True,
+ u'format': '141',
},
},
]
@classmethod
def suitable(cls, url):
"""Receives a URL and returns True if suitable for this IE."""
- if YoutubePlaylistIE.suitable(url) or YoutubeSubscriptionsIE.suitable(url): return False
- return re.match(cls._VALID_URL, url, re.VERBOSE) is not None
+ if YoutubePlaylistIE.suitable(url): return False
+ return re.match(cls._VALID_URL, url) is not None
- def report_video_webpage_download(self, video_id):
- """Report attempt to download video webpage."""
- self.to_screen(u'%s: Downloading video webpage' % video_id)
+ def __init__(self, *args, **kwargs):
+ super(YoutubeIE, self).__init__(*args, **kwargs)
+ self._player_cache = {}
def report_video_info_webpage_download(self, video_id):
"""Report attempt to download video info webpage."""
self.to_screen(u'%s: Downloading video info webpage' % video_id)
- def report_video_subtitles_download(self, video_id):
- """Report attempt to download video info webpage."""
- self.to_screen(u'%s: Checking available subtitles' % video_id)
-
- def report_video_subtitles_request(self, video_id, sub_lang, format):
- """Report attempt to download video info webpage."""
- self.to_screen(u'%s: Downloading video subtitles for %s.%s' % (video_id, sub_lang, format))
-
- def report_video_subtitles_available(self, video_id, sub_lang_list):
- """Report available subtitles."""
- sub_lang = ",".join(list(sub_lang_list.keys()))
- self.to_screen(u'%s: Available subtitles for video: %s' % (video_id, sub_lang))
-
def report_information_extraction(self, video_id):
"""Report attempt to extract video information."""
self.to_screen(u'%s: Extracting video information' % video_id)
"""Indicate the download will use the RTMP protocol."""
self.to_screen(u'RTMP download detected')
- def _decrypt_signature(self, s):
- """Turn the encrypted s field into a working signature"""
-
- if len(s) == 92:
- return s[25] + s[3:25] + s[0] + s[26:42] + s[79] + s[43:79] + s[91] + s[80:83]
- elif len(s) == 90:
- return s[25] + s[3:25] + s[2] + s[26:40] + s[77] + s[41:77] + s[89] + s[78:81]
- elif len(s) == 88:
- return s[48] + s[81:67:-1] + s[82] + s[66:62:-1] + s[85] + s[61:48:-1] + s[67] + s[47:12:-1] + s[3] + s[11:3:-1] + s[2] + s[12]
- elif len(s) == 87:
- return s[83:53:-1] + s[3] + s[52:40:-1] + s[86] + s[39:10:-1] + s[0] + s[9:3:-1] + s[53]
- elif len(s) == 86:
- return s[83:85] + s[26] + s[79:46:-1] + s[85] + s[45:36:-1] + s[30] + s[35:30:-1] + s[46] + s[29:26:-1] + s[82] + s[25:1:-1]
- elif len(s) == 85:
- return s[2:8] + s[0] + s[9:21] + s[65] + s[22:65] + s[84] + s[66:82] + s[21]
- elif len(s) == 84:
- return s[83:27:-1] + s[0] + s[26:5:-1] + s[2:0:-1] + s[27]
- elif len(s) == 83:
- return s[:15] + s[80] + s[16:80] + s[15]
- elif len(s) == 82:
- return s[36] + s[79:67:-1] + s[81] + s[66:40:-1] + s[33] + s[39:36:-1] + s[40] + s[35] + s[0] + s[67] + s[32:0:-1] + s[34]
- elif len(s) == 81:
- return s[56] + s[79:56:-1] + s[41] + s[55:41:-1] + s[80] + s[40:34:-1] + s[0] + s[33:29:-1] + s[34] + s[28:9:-1] + s[29] + s[8:0:-1] + s[9]
- elif len(s) == 79:
- return s[54] + s[77:54:-1] + s[39] + s[53:39:-1] + s[78] + s[38:34:-1] + s[0] + s[33:29:-1] + s[34] + s[28:9:-1] + s[29] + s[8:0:-1] + s[9]
-
+ def _extract_signature_function(self, video_id, player_url, slen):
+ id_m = re.match(r'.*-(?P<id>[a-zA-Z0-9_-]+)\.(?P<ext>[a-z]+)$',
+ player_url)
+ player_type = id_m.group('ext')
+ player_id = id_m.group('id')
+
+ # Read from filesystem cache
+ func_id = '%s_%s_%d' % (player_type, player_id, slen)
+ assert os.path.basename(func_id) == func_id
+ cache_dir = get_cachedir(self._downloader.params)
+
+ cache_enabled = cache_dir is not None
+ if cache_enabled:
+ cache_fn = os.path.join(os.path.expanduser(cache_dir),
+ u'youtube-sigfuncs',
+ func_id + '.json')
+ try:
+ with io.open(cache_fn, 'r', encoding='utf-8') as cachef:
+ cache_spec = json.load(cachef)
+ return lambda s: u''.join(s[i] for i in cache_spec)
+ except IOError:
+ pass # No cache available
+
+ if player_type == 'js':
+ code = self._download_webpage(
+ player_url, video_id,
+ note=u'Downloading %s player %s' % (player_type, player_id),
+ errnote=u'Download of %s failed' % player_url)
+ res = self._parse_sig_js(code)
+ elif player_type == 'swf':
+ urlh = self._request_webpage(
+ player_url, video_id,
+ note=u'Downloading %s player %s' % (player_type, player_id),
+ errnote=u'Download of %s failed' % player_url)
+ code = urlh.read()
+ res = self._parse_sig_swf(code)
else:
- raise ExtractorError(u'Unable to decrypt signature, key length %d not supported; retrying might work' % (len(s)))
+ assert False, 'Invalid player type %r' % player_type
- def _decrypt_signature_age_gate(self, s):
- # The videos with age protection use another player, so the algorithms
- # can be different.
- if len(s) == 86:
- return s[2:63] + s[82] + s[64:82] + s[63]
+ if cache_enabled:
+ try:
+ test_string = u''.join(map(compat_chr, range(slen)))
+ cache_res = res(test_string)
+ cache_spec = [ord(c) for c in cache_res]
+ try:
+ os.makedirs(os.path.dirname(cache_fn))
+ except OSError as ose:
+ if ose.errno != errno.EEXIST:
+ raise
+ write_json_file(cache_spec, cache_fn)
+ except Exception:
+ tb = traceback.format_exc()
+ self._downloader.report_warning(
+ u'Writing cache to %r failed: %s' % (cache_fn, tb))
+
+ return res
+
+ def _print_sig_code(self, func, slen):
+ def gen_sig_code(idxs):
+ def _genslice(start, end, step):
+ starts = u'' if start == 0 else str(start)
+ ends = (u':%d' % (end+step)) if end + step >= 0 else u':'
+ steps = u'' if step == 1 else (u':%d' % step)
+ return u's[%s%s%s]' % (starts, ends, steps)
+
+ step = None
+ start = '(Never used)' # Quelch pyflakes warnings - start will be
+ # set as soon as step is set
+ for i, prev in zip(idxs[1:], idxs[:-1]):
+ if step is not None:
+ if i - prev == step:
+ continue
+ yield _genslice(start, prev, step)
+ step = None
+ continue
+ if i - prev in [-1, 1]:
+ step = i - prev
+ start = prev
+ continue
+ else:
+ yield u's[%d]' % prev
+ if step is None:
+ yield u's[%d]' % i
+ else:
+ yield _genslice(start, i, step)
+
+ test_string = u''.join(map(compat_chr, range(slen)))
+ cache_res = func(test_string)
+ cache_spec = [ord(c) for c in cache_res]
+ expr_code = u' + '.join(gen_sig_code(cache_spec))
+ code = u'if len(s) == %d:\n return %s\n' % (slen, expr_code)
+ self.to_screen(u'Extracted signature function:\n' + code)
+
+ def _parse_sig_js(self, jscode):
+ funcname = self._search_regex(
+ r'signature=([$a-zA-Z]+)', jscode,
+ u'Initial JS player signature function name')
+
+ jsi = JSInterpreter(jscode)
+ initial_function = jsi.extract_function(funcname)
+ return lambda s: initial_function([s])
+
+ def _parse_sig_swf(self, file_contents):
+ if file_contents[1:3] != b'WS':
+ raise ExtractorError(
+ u'Not an SWF file; header is %r' % file_contents[:3])
+ if file_contents[:1] == b'C':
+ content = zlib.decompress(file_contents[8:])
else:
- # Fallback to the other algortihms
- return self._decrypt_signature(s)
+ raise NotImplementedError(u'Unsupported compression format %r' %
+ file_contents[:1])
+
+ def extract_tags(content):
+ pos = 0
+ while pos < len(content):
+ header16 = struct.unpack('<H', content[pos:pos+2])[0]
+ pos += 2
+ tag_code = header16 >> 6
+ tag_len = header16 & 0x3f
+ if tag_len == 0x3f:
+ tag_len = struct.unpack('<I', content[pos:pos+4])[0]
+ pos += 4
+ assert pos+tag_len <= len(content)
+ yield (tag_code, content[pos:pos+tag_len])
+ pos += tag_len
+
+ code_tag = next(tag
+ for tag_code, tag in extract_tags(content)
+ if tag_code == 82)
+ p = code_tag.index(b'\0', 4) + 1
+ code_reader = io.BytesIO(code_tag[p:])
+
+ # Parse ABC (AVM2 ByteCode)
+ def read_int(reader=None):
+ if reader is None:
+ reader = code_reader
+ res = 0
+ shift = 0
+ for _ in range(5):
+ buf = reader.read(1)
+ assert len(buf) == 1
+ b = struct.unpack('<B', buf)[0]
+ res = res | ((b & 0x7f) << shift)
+ if b & 0x80 == 0:
+ break
+ shift += 7
+ return res
+
+ def u30(reader=None):
+ res = read_int(reader)
+ assert res & 0xf0000000 == 0
+ return res
+ u32 = read_int
+
+ def s32(reader=None):
+ v = read_int(reader)
+ if v & 0x80000000 != 0:
+ v = - ((v ^ 0xffffffff) + 1)
+ return v
+
+ def read_string(reader=None):
+ if reader is None:
+ reader = code_reader
+ slen = u30(reader)
+ resb = reader.read(slen)
+ assert len(resb) == slen
+ return resb.decode('utf-8')
+
+ def read_bytes(count, reader=None):
+ if reader is None:
+ reader = code_reader
+ resb = reader.read(count)
+ assert len(resb) == count
+ return resb
+
+ def read_byte(reader=None):
+ resb = read_bytes(1, reader=reader)
+ res = struct.unpack('<B', resb)[0]
+ return res
+
+ # minor_version + major_version
+ read_bytes(2 + 2)
+
+ # Constant pool
+ int_count = u30()
+ for _c in range(1, int_count):
+ s32()
+ uint_count = u30()
+ for _c in range(1, uint_count):
+ u32()
+ double_count = u30()
+ read_bytes((double_count-1) * 8)
+ string_count = u30()
+ constant_strings = [u'']
+ for _c in range(1, string_count):
+ s = read_string()
+ constant_strings.append(s)
+ namespace_count = u30()
+ for _c in range(1, namespace_count):
+ read_bytes(1) # kind
+ u30() # name
+ ns_set_count = u30()
+ for _c in range(1, ns_set_count):
+ count = u30()
+ for _c2 in range(count):
+ u30()
+ multiname_count = u30()
+ MULTINAME_SIZES = {
+ 0x07: 2, # QName
+ 0x0d: 2, # QNameA
+ 0x0f: 1, # RTQName
+ 0x10: 1, # RTQNameA
+ 0x11: 0, # RTQNameL
+ 0x12: 0, # RTQNameLA
+ 0x09: 2, # Multiname
+ 0x0e: 2, # MultinameA
+ 0x1b: 1, # MultinameL
+ 0x1c: 1, # MultinameLA
+ }
+ multinames = [u'']
+ for _c in range(1, multiname_count):
+ kind = u30()
+ assert kind in MULTINAME_SIZES, u'Invalid multiname kind %r' % kind
+ if kind == 0x07:
+ u30() # namespace_idx
+ name_idx = u30()
+ multinames.append(constant_strings[name_idx])
+ else:
+ multinames.append('[MULTINAME kind: %d]' % kind)
+ for _c2 in range(MULTINAME_SIZES[kind]):
+ u30()
+
+ # Methods
+ method_count = u30()
+ MethodInfo = collections.namedtuple(
+ 'MethodInfo',
+ ['NEED_ARGUMENTS', 'NEED_REST'])
+ method_infos = []
+ for method_id in range(method_count):
+ param_count = u30()
+ u30() # return type
+ for _ in range(param_count):
+ u30() # param type
+ u30() # name index (always 0 for youtube)
+ flags = read_byte()
+ if flags & 0x08 != 0:
+ # Options present
+ option_count = u30()
+ for c in range(option_count):
+ u30() # val
+ read_bytes(1) # kind
+ if flags & 0x80 != 0:
+ # Param names present
+ for _ in range(param_count):
+ u30() # param name
+ mi = MethodInfo(flags & 0x01 != 0, flags & 0x04 != 0)
+ method_infos.append(mi)
+
+ # Metadata
+ metadata_count = u30()
+ for _c in range(metadata_count):
+ u30() # name
+ item_count = u30()
+ for _c2 in range(item_count):
+ u30() # key
+ u30() # value
+
+ def parse_traits_info():
+ trait_name_idx = u30()
+ kind_full = read_byte()
+ kind = kind_full & 0x0f
+ attrs = kind_full >> 4
+ methods = {}
+ if kind in [0x00, 0x06]: # Slot or Const
+ u30() # Slot id
+ u30() # type_name_idx
+ vindex = u30()
+ if vindex != 0:
+ read_byte() # vkind
+ elif kind in [0x01, 0x02, 0x03]: # Method / Getter / Setter
+ u30() # disp_id
+ method_idx = u30()
+ methods[multinames[trait_name_idx]] = method_idx
+ elif kind == 0x04: # Class
+ u30() # slot_id
+ u30() # classi
+ elif kind == 0x05: # Function
+ u30() # slot_id
+ function_idx = u30()
+ methods[function_idx] = multinames[trait_name_idx]
+ else:
+ raise ExtractorError(u'Unsupported trait kind %d' % kind)
+
+ if attrs & 0x4 != 0: # Metadata present
+ metadata_count = u30()
+ for _c3 in range(metadata_count):
+ u30() # metadata index
+
+ return methods
+
+ # Classes
+ TARGET_CLASSNAME = u'SignatureDecipher'
+ searched_idx = multinames.index(TARGET_CLASSNAME)
+ searched_class_id = None
+ class_count = u30()
+ for class_id in range(class_count):
+ name_idx = u30()
+ if name_idx == searched_idx:
+ # We found the class we're looking for!
+ searched_class_id = class_id
+ u30() # super_name idx
+ flags = read_byte()
+ if flags & 0x08 != 0: # Protected namespace is present
+ u30() # protected_ns_idx
+ intrf_count = u30()
+ for _c2 in range(intrf_count):
+ u30()
+ u30() # iinit
+ trait_count = u30()
+ for _c2 in range(trait_count):
+ parse_traits_info()
+
+ if searched_class_id is None:
+ raise ExtractorError(u'Target class %r not found' %
+ TARGET_CLASSNAME)
+
+ method_names = {}
+ method_idxs = {}
+ for class_id in range(class_count):
+ u30() # cinit
+ trait_count = u30()
+ for _c2 in range(trait_count):
+ trait_methods = parse_traits_info()
+ if class_id == searched_class_id:
+ method_names.update(trait_methods.items())
+ method_idxs.update(dict(
+ (idx, name)
+ for name, idx in trait_methods.items()))
+
+ # Scripts
+ script_count = u30()
+ for _c in range(script_count):
+ u30() # init
+ trait_count = u30()
+ for _c2 in range(trait_count):
+ parse_traits_info()
+
+ # Method bodies
+ method_body_count = u30()
+ Method = collections.namedtuple('Method', ['code', 'local_count'])
+ methods = {}
+ for _c in range(method_body_count):
+ method_idx = u30()
+ u30() # max_stack
+ local_count = u30()
+ u30() # init_scope_depth
+ u30() # max_scope_depth
+ code_length = u30()
+ code = read_bytes(code_length)
+ if method_idx in method_idxs:
+ m = Method(code, local_count)
+ methods[method_idxs[method_idx]] = m
+ exception_count = u30()
+ for _c2 in range(exception_count):
+ u30() # from
+ u30() # to
+ u30() # target
+ u30() # exc_type
+ u30() # var_name
+ trait_count = u30()
+ for _c2 in range(trait_count):
+ parse_traits_info()
+
+ assert p + code_reader.tell() == len(code_tag)
+ assert len(methods) == len(method_idxs)
+
+ method_pyfunctions = {}
+
+ def extract_function(func_name):
+ if func_name in method_pyfunctions:
+ return method_pyfunctions[func_name]
+ if func_name not in methods:
+ raise ExtractorError(u'Cannot find function %r' % func_name)
+ m = methods[func_name]
+
+ def resfunc(args):
+ registers = ['(this)'] + list(args) + [None] * m.local_count
+ stack = []
+ coder = io.BytesIO(m.code)
+ while True:
+ opcode = struct.unpack('!B', coder.read(1))[0]
+ if opcode == 36: # pushbyte
+ v = struct.unpack('!B', coder.read(1))[0]
+ stack.append(v)
+ elif opcode == 44: # pushstring
+ idx = u30(coder)
+ stack.append(constant_strings[idx])
+ elif opcode == 48: # pushscope
+ # We don't implement the scope register, so we'll just
+ # ignore the popped value
+ stack.pop()
+ elif opcode == 70: # callproperty
+ index = u30(coder)
+ mname = multinames[index]
+ arg_count = u30(coder)
+ args = list(reversed(
+ [stack.pop() for _ in range(arg_count)]))
+ obj = stack.pop()
+ if mname == u'split':
+ assert len(args) == 1
+ assert isinstance(args[0], compat_str)
+ assert isinstance(obj, compat_str)
+ if args[0] == u'':
+ res = list(obj)
+ else:
+ res = obj.split(args[0])
+ stack.append(res)
+ elif mname == u'slice':
+ assert len(args) == 1
+ assert isinstance(args[0], int)
+ assert isinstance(obj, list)
+ res = obj[args[0]:]
+ stack.append(res)
+ elif mname == u'join':
+ assert len(args) == 1
+ assert isinstance(args[0], compat_str)
+ assert isinstance(obj, list)
+ res = args[0].join(obj)
+ stack.append(res)
+ elif mname in method_pyfunctions:
+ stack.append(method_pyfunctions[mname](args))
+ else:
+ raise NotImplementedError(
+ u'Unsupported property %r on %r'
+ % (mname, obj))
+ elif opcode == 72: # returnvalue
+ res = stack.pop()
+ return res
+ elif opcode == 79: # callpropvoid
+ index = u30(coder)
+ mname = multinames[index]
+ arg_count = u30(coder)
+ args = list(reversed(
+ [stack.pop() for _ in range(arg_count)]))
+ obj = stack.pop()
+ if mname == u'reverse':
+ assert isinstance(obj, list)
+ obj.reverse()
+ else:
+ raise NotImplementedError(
+ u'Unsupported (void) property %r on %r'
+ % (mname, obj))
+ elif opcode == 93: # findpropstrict
+ index = u30(coder)
+ mname = multinames[index]
+ res = extract_function(mname)
+ stack.append(res)
+ elif opcode == 97: # setproperty
+ index = u30(coder)
+ value = stack.pop()
+ idx = stack.pop()
+ obj = stack.pop()
+ assert isinstance(obj, list)
+ assert isinstance(idx, int)
+ obj[idx] = value
+ elif opcode == 98: # getlocal
+ index = u30(coder)
+ stack.append(registers[index])
+ elif opcode == 99: # setlocal
+ index = u30(coder)
+ value = stack.pop()
+ registers[index] = value
+ elif opcode == 102: # getproperty
+ index = u30(coder)
+ pname = multinames[index]
+ if pname == u'length':
+ obj = stack.pop()
+ assert isinstance(obj, list)
+ stack.append(len(obj))
+ else: # Assume attribute access
+ idx = stack.pop()
+ assert isinstance(idx, int)
+ obj = stack.pop()
+ assert isinstance(obj, list)
+ stack.append(obj[idx])
+ elif opcode == 128: # coerce
+ u30(coder)
+ elif opcode == 133: # coerce_s
+ assert isinstance(stack[-1], (type(None), compat_str))
+ elif opcode == 164: # modulo
+ value2 = stack.pop()
+ value1 = stack.pop()
+ res = value1 % value2
+ stack.append(res)
+ elif opcode == 208: # getlocal_0
+ stack.append(registers[0])
+ elif opcode == 209: # getlocal_1
+ stack.append(registers[1])
+ elif opcode == 210: # getlocal_2
+ stack.append(registers[2])
+ elif opcode == 211: # getlocal_3
+ stack.append(registers[3])
+ elif opcode == 214: # setlocal_2
+ registers[2] = stack.pop()
+ elif opcode == 215: # setlocal_3
+ registers[3] = stack.pop()
+ else:
+ raise NotImplementedError(
+ u'Unsupported opcode %d' % opcode)
+
+ method_pyfunctions[func_name] = resfunc
+ return resfunc
+
+ initial_function = extract_function(u'decipher')
+ return lambda s: initial_function([s])
+
+ def _decrypt_signature(self, s, video_id, player_url, age_gate=False):
+ """Turn the encrypted s field into a working signature"""
+ if player_url is None:
+ raise ExtractorError(u'Cannot decrypt signature without player_url')
- def _get_available_subtitles(self, video_id):
- self.report_video_subtitles_download(video_id)
- request = compat_urllib_request.Request('http://video.google.com/timedtext?hl=en&type=list&v=%s' % video_id)
+ if player_url.startswith(u'//'):
+ player_url = u'https:' + player_url
try:
- sub_list = compat_urllib_request.urlopen(request).read().decode('utf-8')
- except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
- return (u'unable to download video subtitles: %s' % compat_str(err), None)
- sub_lang_list = re.findall(r'name="([^"]*)"[^>]+lang_code="([\w\-]+)"', sub_list)
- sub_lang_list = dict((l[1], l[0]) for l in sub_lang_list)
+ player_id = (player_url, len(s))
+ if player_id not in self._player_cache:
+ func = self._extract_signature_function(
+ video_id, player_url, len(s)
+ )
+ self._player_cache[player_id] = func
+ func = self._player_cache[player_id]
+ if self._downloader.params.get('youtube_print_sig_code'):
+ self._print_sig_code(func, len(s))
+ return func(s)
+ except Exception as e:
+ tb = traceback.format_exc()
+ raise ExtractorError(
+ u'Automatic signature extraction failed: ' + tb, cause=e)
+
+ def _get_available_subtitles(self, video_id, webpage):
+ try:
+ sub_list = self._download_webpage(
+ '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(u'unable to download video subtitles: %s' % compat_str(err))
+ return {}
+ lang_list = re.findall(r'name="([^"]*)"[^>]+lang_code="([\w\-]+)"', sub_list)
+
+ sub_lang_list = {}
+ for l in lang_list:
+ lang = l[1]
+ params = compat_urllib_parse.urlencode({
+ 'lang': lang,
+ 'v': video_id,
+ 'fmt': self._downloader.params.get('subtitlesformat', 'srt'),
+ 'name': unescapeHTML(l[0]).encode('utf-8'),
+ })
+ url = u'https://www.youtube.com/api/timedtext?' + params
+ sub_lang_list[lang] = url
if not sub_lang_list:
- return (u'video doesn\'t have subtitles', None)
+ self._downloader.report_warning(u'video doesn\'t have subtitles')
+ return {}
return sub_lang_list
- def _list_available_subtitles(self, video_id):
- sub_lang_list = self._get_available_subtitles(video_id)
- self.report_video_subtitles_available(video_id, sub_lang_list)
-
- def _request_subtitle(self, sub_lang, sub_name, video_id, format):
- """
- Return tuple:
- (error_message, sub_lang, sub)
- """
- self.report_video_subtitles_request(video_id, sub_lang, format)
- params = compat_urllib_parse.urlencode({
- 'lang': sub_lang,
- 'name': sub_name,
- 'v': video_id,
- 'fmt': format,
- })
- url = 'http://www.youtube.com/api/timedtext?' + params
- try:
- sub = compat_urllib_request.urlopen(url).read().decode('utf-8')
- except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
- return (u'unable to download video subtitles: %s' % compat_str(err), None, None)
- if not sub:
- return (u'Did not fetch video subtitles', None, None)
- return (None, sub_lang, sub)
-
- def _request_automatic_caption(self, video_id, webpage):
+ def _get_available_automatic_caption(self, video_id, webpage):
"""We need the webpage for getting the captions url, pass it as an
argument to speed up the process."""
- sub_lang = self._downloader.params.get('subtitleslang') or 'en'
- sub_format = self._downloader.params.get('subtitlesformat')
+ sub_format = self._downloader.params.get('subtitlesformat', 'srt')
self.to_screen(u'%s: Looking for automatic captions' % video_id)
mobj = re.search(r';ytplayer.config = ({.*?});', webpage)
- err_msg = u'Couldn\'t find automatic captions for "%s"' % sub_lang
+ err_msg = u'Couldn\'t find automatic captions for %s' % video_id
if mobj is None:
- return [(err_msg, None, None)]
+ self._downloader.report_warning(err_msg)
+ return {}
player_config = json.loads(mobj.group(1))
try:
args = player_config[u'args']
caption_url = args[u'ttsurl']
timestamp = args[u'timestamp']
- params = compat_urllib_parse.urlencode({
- 'lang': 'en',
- 'tlang': sub_lang,
- 'fmt': sub_format,
- 'ts': timestamp,
- 'kind': 'asr',
+ # We get the available subtitles
+ list_params = compat_urllib_parse.urlencode({
+ 'type': 'list',
+ 'tlangs': 1,
+ 'asrs': 1,
})
- subtitles_url = caption_url + '&' + params
- sub = self._download_webpage(subtitles_url, video_id, u'Downloading automatic captions')
- return [(None, sub_lang, sub)]
- except KeyError:
- return [(err_msg, None, None)]
-
- def _extract_subtitle(self, video_id):
- """
- Return a list with a tuple:
- [(error_message, sub_lang, sub)]
- """
- sub_lang_list = self._get_available_subtitles(video_id)
- sub_format = self._downloader.params.get('subtitlesformat')
- if isinstance(sub_lang_list,tuple): #There was some error, it didn't get the available subtitles
- return [(sub_lang_list[0], None, None)]
- if self._downloader.params.get('subtitleslang', False):
- sub_lang = self._downloader.params.get('subtitleslang')
- elif 'en' in sub_lang_list:
- sub_lang = 'en'
- else:
- sub_lang = list(sub_lang_list.keys())[0]
- if not sub_lang in sub_lang_list:
- return [(u'no closed captions found in the specified language "%s"' % sub_lang, None, None)]
-
- subtitle = self._request_subtitle(sub_lang, sub_lang_list[sub_lang].encode('utf-8'), video_id, sub_format)
- return [subtitle]
-
- def _extract_all_subtitles(self, video_id):
- sub_lang_list = self._get_available_subtitles(video_id)
- sub_format = self._downloader.params.get('subtitlesformat')
- if isinstance(sub_lang_list,tuple): #There was some error, it didn't get the available subtitles
- return [(sub_lang_list[0], None, None)]
- subtitles = []
- for sub_lang in sub_lang_list:
- subtitle = self._request_subtitle(sub_lang, sub_lang_list[sub_lang].encode('utf-8'), video_id, sub_format)
- subtitles.append(subtitle)
- return subtitles
-
- def _print_formats(self, formats):
- print('Available formats:')
- for x in formats:
- print('%s\t:\t%s\t[%s]%s' %(x, self._video_extensions.get(x, 'flv'),
- self._video_dimensions.get(x, '???'),
- ' (3D)' if x in self._3d_itags else ''))
-
- def _extract_id(self, url):
- mobj = re.match(self._VALID_URL, url, re.VERBOSE)
+ 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 or original_lang_node.attrib.get('kind') != 'asr' :
+ self._downloader.report_warning(u'Video doesn\'t have automatic captions')
+ return {}
+ original_lang = original_lang_node.attrib['lang_code']
+
+ sub_lang_list = {}
+ for lang_node in caption_list.findall('target'):
+ sub_lang = lang_node.attrib['lang_code']
+ params = compat_urllib_parse.urlencode({
+ 'lang': original_lang,
+ 'tlang': sub_lang,
+ 'fmt': sub_format,
+ 'ts': timestamp,
+ 'kind': 'asr',
+ })
+ sub_lang_list[sub_lang] = caption_url + '&' + params
+ return 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):
+ self._downloader.report_warning(err_msg)
+ return {}
+
+ @classmethod
+ def extract_id(cls, url):
+ mobj = re.match(cls._VALID_URL, url, re.VERBOSE)
if mobj is None:
raise ExtractorError(u'Invalid URL: %s' % url)
video_id = mobj.group(2)
return video_id
- def _get_video_url_list(self, url_map):
- """
- Transform a dictionary in the format {itag:url} to a list of (itag, url)
- with the requested formats.
- """
- req_format = self._downloader.params.get('format', None)
- format_limit = self._downloader.params.get('format_limit', None)
- available_formats = self._available_formats_prefer_free if self._downloader.params.get('prefer_free_formats', False) else self._available_formats
- if format_limit is not None and format_limit in available_formats:
- format_list = available_formats[available_formats.index(format_limit):]
- else:
- format_list = available_formats
- existing_formats = [x for x in format_list if x in url_map]
- if len(existing_formats) == 0:
- raise ExtractorError(u'no known formats available for video')
- if self._downloader.params.get('listformats', None):
- self._print_formats(existing_formats)
- return
- if req_format is None or req_format == 'best':
- video_url_list = [(existing_formats[0], url_map[existing_formats[0]])] # Best quality
- elif req_format == 'worst':
- video_url_list = [(existing_formats[-1], url_map[existing_formats[-1]])] # worst quality
- elif req_format in ('-1', 'all'):
- video_url_list = [(f, url_map[f]) for f in existing_formats] # All formats
- else:
- # Specific formats. We pick the first in a slash-delimeted sequence.
- # For example, if '1/2/3/4' is requested and '2' and '4' are available, we pick '2'.
- req_formats = req_format.split('/')
- video_url_list = None
- for rf in req_formats:
- if rf in url_map:
- video_url_list = [(rf, url_map[rf])]
- break
- if video_url_list is None:
- raise ExtractorError(u'requested format not available')
- return video_url_list
-
def _extract_from_m3u8(self, manifest_url, video_id):
url_map = {}
def _get_urls(_manifest):
url_map[itag] = format_url
return url_map
+ def _extract_annotations(self, video_id):
+ url = 'https://www.youtube.com/annotations_invideo?features=1&legacy=1&video_id=%s' % video_id
+ return self._download_webpage(url, video_id, note=u'Searching for annotations.', errnote=u'Unable to download video annotations.')
+
def _real_extract(self, url):
- if re.match(r'(?:https?://)?[^/]+/watch\?feature=[a-z_]+$', url):
- self._downloader.report_warning(u'Did you forget to quote the URL? Remember that & is a meta-character in most shells, so you want to put the URL in quotes, like youtube-dl \'http://www.youtube.com/watch?feature=foo&v=BaW_jenozKc\' (or simply youtube-dl BaW_jenozKc ).')
+ proto = (
+ u'http' if self._downloader.params.get('prefer_insecure', False)
+ else u'https')
# 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 = 'https://www.youtube.com/' + compat_urllib_parse.unquote(mobj.group(1)).lstrip('/')
- video_id = self._extract_id(url)
+ url = proto + '://www.youtube.com/' + compat_urllib_parse.unquote(mobj.group(1)).lstrip('/')
+ video_id = self.extract_id(url)
# Get video webpage
- self.report_video_webpage_download(video_id)
- url = 'https://www.youtube.com/watch?v=%s&gl=US&hl=en&has_verified=1' % video_id
- request = compat_urllib_request.Request(url)
- try:
- video_webpage_bytes = compat_urllib_request.urlopen(request).read()
- except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
- raise ExtractorError(u'Unable to download video webpage: %s' % compat_str(err))
-
- video_webpage = video_webpage_bytes.decode('utf-8', 'ignore')
+ url = proto + '://www.youtube.com/watch?v=%s&gl=US&hl=en&has_verified=1' % video_id
+ video_webpage = self._download_webpage(url, video_id)
# Attempt to extract SWF player URL
- mobj = re.search(r'swfConfig.*?"(http:\\/\\/.*?watch.*?-.*?\.swf)"', video_webpage)
+ 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:
# We simulate the access to the video from www.youtube.com/v/{video_id}
# this can be viewed without login into Youtube
data = compat_urllib_parse.urlencode({'video_id': video_id,
- 'el': 'embedded',
+ 'el': 'player_embedded',
'gl': 'US',
'hl': 'en',
'eurl': 'https://youtube.googleapis.com/v/' + video_id,
'asv': 3,
'sts':'1588',
})
- video_info_url = 'https://www.youtube.com/get_video_info?' + data
+ video_info_url = proto + '://www.youtube.com/get_video_info?' + data
video_info_webpage = self._download_webpage(video_info_url, video_id,
note=False,
errnote='unable to download video info webpage')
else:
age_gate = False
for el_type in ['&el=embedded', '&el=detailpage', '&el=vevo', '']:
- video_info_url = ('https://www.youtube.com/get_video_info?&video_id=%s%s&ps=default&eurl=&gl=US&hl=en'
+ video_info_url = (proto + '://www.youtube.com/get_video_info?&video_id=%s%s&ps=default&eurl=&gl=US&hl=en'
% (video_id, el_type))
video_info_webpage = self._download_webpage(video_info_url, video_id,
note=False,
break
if 'token' not in video_info:
if 'reason' in video_info:
- raise ExtractorError(u'YouTube said: %s' % video_info['reason'][0], expected=True)
+ raise ExtractorError(
+ u'YouTube said: %s' % video_info['reason'][0],
+ expected=True, video_id=video_id)
else:
- raise ExtractorError(u'"token" parameter not in video info for unknown reason')
+ raise ExtractorError(
+ u'"token" parameter not in video info for unknown reason',
+ video_id=video_id)
+
+ if 'view_count' in video_info:
+ view_count = int(video_info['view_count'][0])
+ else:
+ view_count = None
# Check for "rental" videos
if 'ypc_video_rental_bar_text' in video_info and 'author' not in video_info:
self._downloader.report_warning(u'unable to extract uploader nickname')
# title
- if 'title' not in video_info:
- raise ExtractorError(u'Unable to extract video title')
- video_title = compat_urllib_parse.unquote_plus(video_info['title'][0])
+ if 'title' in video_info:
+ video_title = video_info['title'][0]
+ else:
+ self._downloader.report_warning(u'Unable to extract video title')
+ video_title = u'_'
# thumbnail image
# We try first to get a high quality image:
video_thumbnail = m_thumb.group(1)
elif 'thumbnail_url' not in video_info:
self._downloader.report_warning(u'unable to extract video thumbnail')
- video_thumbnail = ''
+ video_thumbnail = None
else: # don't panic if we can't find it
video_thumbnail = compat_urllib_parse.unquote_plus(video_info['thumbnail_url'][0])
# upload date
upload_date = None
- mobj = re.search(r'id="eow-date.*?>(.*?)</span>', video_webpage, re.DOTALL)
+ mobj = re.search(r'(?s)id="eow-date.*?>(.*?)</span>', video_webpage)
+ if mobj is None:
+ mobj = re.search(
+ r'(?s)id="watch-uploader-info".*?>.*?(?:Published|Uploaded|Streamed live) on (.*?)</strong>',
+ video_webpage)
if mobj is not None:
upload_date = ' '.join(re.sub(r'[/,-]', r' ', mobj.group(1)).split())
upload_date = unified_strdate(upload_date)
+ m_cat_container = get_element_by_id("eow-category", video_webpage)
+ if m_cat_container:
+ category = self._html_search_regex(
+ r'(?s)<a[^<]+>(.*?)</a>', m_cat_container, 'category',
+ default=None)
+ video_categories = None if category is None else [category]
+ else:
+ video_categories = None
+
# 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)
else:
video_description = u''
+ def _extract_count(klass):
+ count = self._search_regex(
+ r'class="%s">([\d,]+)</span>' % re.escape(klass),
+ video_webpage, klass, default=None)
+ if count is not None:
+ return int(count.replace(',', ''))
+ return None
+ like_count = _extract_count(u'likes-count')
+ dislike_count = _extract_count(u'dislikes-count')
+
# subtitles
- video_subtitles = None
-
- if self._downloader.params.get('writesubtitles', False):
- video_subtitles = self._extract_subtitle(video_id)
- if video_subtitles:
- (sub_error, sub_lang, sub) = video_subtitles[0]
- if sub_error:
- self._downloader.report_warning(sub_error)
-
- if self._downloader.params.get('writeautomaticsub', False):
- video_subtitles = self._request_automatic_caption(video_id, video_webpage)
- (sub_error, sub_lang, sub) = video_subtitles[0]
- if sub_error:
- self._downloader.report_warning(sub_error)
-
- if self._downloader.params.get('allsubtitles', False):
- video_subtitles = self._extract_all_subtitles(video_id)
- for video_subtitle in video_subtitles:
- (sub_error, sub_lang, sub) = video_subtitle
- if sub_error:
- self._downloader.report_warning(sub_error)
+ video_subtitles = self.extract_subtitles(video_id, video_webpage)
if self._downloader.params.get('listsubtitles', False):
- self._list_available_subtitles(video_id)
+ self._list_available_subtitles(video_id, video_webpage)
return
if 'length_seconds' not in video_info:
self._downloader.report_warning(u'unable to extract video duration')
- video_duration = ''
+ video_duration = None
else:
- video_duration = compat_urllib_parse.unquote_plus(video_info['length_seconds'][0])
+ video_duration = int(compat_urllib_parse.unquote_plus(video_info['length_seconds'][0]))
- # Decide which formats to download
+ # annotations
+ video_annotations = None
+ if self._downloader.params.get('writeannotations', False):
+ video_annotations = self._extract_annotations(video_id)
+ # Decide which formats to download
try:
- mobj = re.search(r';ytplayer.config = ({.*?});', video_webpage)
+ mobj = re.search(r';ytplayer\.config\s*=\s*({.*?});', video_webpage)
if not mobj:
raise ValueError('Could not find vevo ID')
- info = json.loads(mobj.group(1))
- args = info['args']
+ json_code = uppercase_escape(mobj.group(1))
+ ytplayer_config = json.loads(json_code)
+ args = ytplayer_config['args']
# Easy way to know if the 's' value is in url_encoded_fmt_stream_map
# this signatures are encrypted
- m_s = re.search(r'[&,]s=', args['url_encoded_fmt_stream_map'])
+ if 'url_encoded_fmt_stream_map' not in args:
+ raise ValueError(u'No stream_map present') # caught below
+ re_signature = re.compile(r'[&,]s=')
+ m_s = re_signature.search(args['url_encoded_fmt_stream_map'])
if m_s is not None:
self.to_screen(u'%s: Encrypted signatures detected.' % video_id)
video_info['url_encoded_fmt_stream_map'] = [args['url_encoded_fmt_stream_map']]
+ m_s = re_signature.search(args.get('adaptive_fmts', u''))
+ if m_s is not None:
+ if 'adaptive_fmts' in video_info:
+ video_info['adaptive_fmts'][0] += ',' + args['adaptive_fmts']
+ else:
+ video_info['adaptive_fmts'] = [args['adaptive_fmts']]
except ValueError:
pass
+ def _map_to_format_list(urlmap):
+ formats = []
+ for itag, video_real_url in urlmap.items():
+ dct = {
+ 'format_id': itag,
+ 'url': video_real_url,
+ 'player_url': player_url,
+ }
+ if itag in self._formats:
+ dct.update(self._formats[itag])
+ formats.append(dct)
+ return formats
+
if 'conn' in video_info and video_info['conn'][0].startswith('rtmp'):
self.report_rtmp_download()
- video_url_list = [(None, video_info['conn'][0])]
- elif 'url_encoded_fmt_stream_map' in video_info and len(video_info['url_encoded_fmt_stream_map']) >= 1:
- if 'rtmpe%3Dyes' in video_info['url_encoded_fmt_stream_map'][0]:
+ formats = [{
+ 'format_id': '_rtmp',
+ 'protocol': 'rtmp',
+ 'url': video_info['conn'][0],
+ 'player_url': player_url,
+ }]
+ elif len(video_info.get('url_encoded_fmt_stream_map', [])) >= 1 or len(video_info.get('adaptive_fmts', [])) >= 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)
url_map = {}
- for url_data_str in video_info['url_encoded_fmt_stream_map'][0].split(','):
+ for url_data_str in encoded_url_map.split(','):
url_data = compat_parse_qs(url_data_str)
if 'itag' in url_data and 'url' in url_data:
url = url_data['url'][0]
if 'sig' in url_data:
url += '&signature=' + url_data['sig'][0]
elif 's' in url_data:
+ encrypted_sig = url_data['s'][0]
if self._downloader.params.get('verbose'):
- s = url_data['s'][0]
if age_gate:
- player_version = self._search_regex(r'ad3-(.+?)\.swf',
- video_info['ad3_module'][0] if 'ad3_module' in video_info else 'NOT FOUND',
- 'flash player', fatal=False)
- player = 'flash player %s' % player_version
+ if player_url is None:
+ player_version = 'unknown'
+ else:
+ player_version = self._search_regex(
+ r'-(.+)\.swf$', player_url,
+ u'flash player', fatal=False)
+ player_desc = 'flash player %s' % player_version
else:
- player = u'html5 player %s' % self._search_regex(r'html5player-(.+?)\.js', video_webpage,
+ player_version = self._search_regex(
+ r'html5player-(.+?)\.js', video_webpage,
'html5 player', fatal=False)
- parts_sizes = u'.'.join(compat_str(len(part)) for part in s.split('.'))
+ player_desc = u'html5 player %s' % player_version
+
+ parts_sizes = u'.'.join(compat_str(len(part)) for part in encrypted_sig.split('.'))
self.to_screen(u'encrypted signature length %d (%s), itag %s, %s' %
- (len(s), parts_sizes, url_data['itag'][0], player))
- encrypted_sig = url_data['s'][0]
- if age_gate:
- signature = self._decrypt_signature_age_gate(encrypted_sig)
- else:
- signature = self._decrypt_signature(encrypted_sig)
+ (len(encrypted_sig), parts_sizes, url_data['itag'][0], player_desc))
+
+ if not age_gate:
+ jsplayer_url_json = self._search_regex(
+ r'"assets":.+?"js":\s*("[^"]+")',
+ video_webpage, u'JS player URL')
+ player_url = json.loads(jsplayer_url_json)
+
+ signature = self._decrypt_signature(
+ encrypted_sig, video_id, player_url, age_gate)
url += '&signature=' + signature
if 'ratebypass' not in url:
url += '&ratebypass=yes'
url_map[url_data['itag'][0]] = url
- video_url_list = self._get_video_url_list(url_map)
- if not video_url_list:
- return
+ formats = _map_to_format_list(url_map)
elif video_info.get('hlsvp'):
manifest_url = video_info['hlsvp'][0]
url_map = self._extract_from_m3u8(manifest_url, video_id)
- video_url_list = self._get_video_url_list(url_map)
- if not video_url_list:
- return
-
+ formats = _map_to_format_list(url_map)
else:
- raise ExtractorError(u'no conn or url_encoded_fmt_stream_map information found in video info')
-
- results = []
- for format_param, video_real_url in video_url_list:
- # Extension
- video_extension = self._video_extensions.get(format_param, 'flv')
-
- video_format = '{0} - {1}{2}'.format(format_param if format_param else video_extension,
- self._video_dimensions.get(format_param, '???'),
- ' (3D)' if format_param in self._3d_itags else '')
-
- results.append({
- 'id': video_id,
- 'url': video_real_url,
- 'uploader': video_uploader,
- 'uploader_id': video_uploader_id,
- 'upload_date': upload_date,
- 'title': video_title,
- 'ext': video_extension,
- 'format': video_format,
- 'thumbnail': video_thumbnail,
- 'description': video_description,
- 'player_url': player_url,
- 'subtitles': video_subtitles,
- 'duration': video_duration
- })
- return results
+ raise ExtractorError(u'no conn, hlsvp or url_encoded_fmt_stream_map information found in video info')
+
+ # Look for the DASH manifest
+ if (self._downloader.params.get('youtube_include_dash_manifest', False)):
+ try:
+ # The DASH manifest used needs to be the one from the original video_webpage.
+ # The one found in get_video_info seems to be using different signatures.
+ # However, in the case of an age restriction there won't be any embedded dashmpd in the video_webpage.
+ # Luckily, it seems, this case uses some kind of default signature (len == 86), so the
+ # combination of get_video_info and the _static_decrypt_signature() decryption fallback will work here.
+ if age_gate:
+ dash_manifest_url = video_info.get('dashmpd')[0]
+ else:
+ dash_manifest_url = ytplayer_config['args']['dashmpd']
+ 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
+ dash_manifest_url = re.sub(r'/s/([\w\.]+)', decrypt_sig, dash_manifest_url)
+ dash_doc = self._download_xml(
+ dash_manifest_url, video_id,
+ note=u'Downloading DASH manifest',
+ errnote=u'Could not download DASH manifest')
+ for r in dash_doc.findall(u'.//{urn:mpeg:DASH:schema:MPD:2011}Representation'):
+ url_el = r.find('{urn:mpeg:DASH:schema:MPD:2011}BaseURL')
+ if url_el is None:
+ continue
+ format_id = r.attrib['id']
+ video_url = url_el.text
+ filesize = int_or_none(url_el.attrib.get('{http://youtube.com/yt/2012/10/10}contentLength'))
+ f = {
+ 'format_id': format_id,
+ 'url': video_url,
+ 'width': int_or_none(r.attrib.get('width')),
+ 'tbr': int_or_none(r.attrib.get('bandwidth'), 1000),
+ 'asr': int_or_none(r.attrib.get('audioSamplingRate')),
+ 'filesize': filesize,
+ }
+ try:
+ existing_format = next(
+ fo for fo in formats
+ if fo['format_id'] == format_id)
+ except StopIteration:
+ f.update(self._formats.get(format_id, {}))
+ formats.append(f)
+ else:
+ existing_format.update(f)
+
+ except (ExtractorError, KeyError) as e:
+ self.report_warning(u'Skipping DASH manifest: %s' % e, video_id)
+
+ self._sort_formats(formats)
+
+ return {
+ 'id': video_id,
+ 'uploader': video_uploader,
+ 'uploader_id': video_uploader_id,
+ 'upload_date': upload_date,
+ 'title': video_title,
+ 'thumbnail': video_thumbnail,
+ 'description': video_description,
+ 'categories': video_categories,
+ 'subtitles': video_subtitles,
+ 'duration': video_duration,
+ 'age_limit': 18 if age_gate else 0,
+ 'annotations': video_annotations,
+ 'webpage_url': proto + '://www.youtube.com/watch?v=%s' % video_id,
+ 'view_count': view_count,
+ 'like_count': like_count,
+ 'dislike_count': dislike_count,
+ 'formats': formats,
+ }
-class YoutubePlaylistIE(InfoExtractor):
+class YoutubePlaylistIE(YoutubeBaseInfoExtractor):
IE_DESC = u'YouTube.com playlists'
- _VALID_URL = r"""(?:
+ _VALID_URL = r"""(?x)(?:
(?:https?://)?
(?:\w+\.)?
youtube\.com/
\? (?:.*?&)*? (?:p|a|list)=
| p/
)
- ((?:PL|EC|UU|FL)?[0-9A-Za-z-_]{10,})
+ (
+ (?:PL|LL|EC|UU|FL|RD)?[0-9A-Za-z-_]{10,}
+ # Top tracks, they can also include dots
+ |(?:MC)[\w\.]*
+ )
.*
|
- ((?:PL|EC|UU|FL)[0-9A-Za-z-_]{10,})
+ ((?:PL|LL|EC|UU|FL|RD)[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'
+ _MORE_PAGES_INDICATOR = r'data-link-type="next"'
+ _VIDEO_RE = r'href="\s*/watch\?v=(?P<id>[0-9A-Za-z_-]{11})&[^"]*?index=(?P<index>\d+)'
IE_NAME = u'youtube:playlist'
- @classmethod
- def suitable(cls, url):
- """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[-11:], playlist_id)
+ webpage = self._download_webpage(url, playlist_id, u'Downloading Youtube mix')
+ search_title = lambda class_name: get_element_by_attribute('class', class_name, webpage)
+ title_span = (search_title('playlist-title') or
+ search_title('title long-title') or search_title('title'))
+ title = clean_html(title_span)
+ video_re = r'''(?x)data-video-username=".*?".*?
+ href="/watch\?v=([0-9A-Za-z_-]{11})&[^"]*?list=%s''' % re.escape(playlist_id)
+ ids = orderedSet(re.findall(video_re, webpage, flags=re.DOTALL))
+ 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)
+ mobj = re.match(self._VALID_URL, url)
if mobj is None:
raise ExtractorError(u'Invalid URL: %s' % url)
-
- # Download playlist videos from API
playlist_id = mobj.group(1) or mobj.group(2)
- videos = []
+
+ # Check if it's a video-specific URL
+ query_dict = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
+ if 'v' in query_dict:
+ 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(video_id, 'Youtube', video_id=video_id)
+ else:
+ self.to_screen(u'Downloading playlist %s - add --no-playlist to just download video %s' % (playlist_id, video_id))
+
+ if playlist_id.startswith('RD'):
+ # Mixes require a custom extraction process
+ return self._extract_mix(playlist_id)
+ if playlist_id.startswith('TL'):
+ raise ExtractorError(u'For downloading YouTube.com top lists, use '
+ u'the "yttoplist" keyword, for example "youtube-dl \'yttoplist:music:Top Tracks\'"', expected=True)
+
+ url = self._TEMPLATE_URL % playlist_id
+ page = self._download_webpage(url, playlist_id)
+ more_widget_html = content_html = page
+
+ # Check if the playlist exists or is private
+ if re.search(r'<div class="yt-alert-message">[^<]*?(The|This) playlist (does not exist|is private)[^<]*?</div>', page) is not None:
+ raise ExtractorError(
+ u'The playlist doesn\'t exist or is private, use --username or '
+ '--netrc to access it.',
+ expected=True)
+
+ # 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')
+ matches = re.finditer(self._VIDEO_RE, content_html)
+ # 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)
+
+ mobj = re.search(r'data-uix-load-more-href="/?(?P<more>[^"]+)"', more_widget_html)
+ if not mobj:
break
- url = self._TEMPLATE_URL % (playlist_id, self._MAX_RESULTS, start_index)
- page = self._download_webpage(url, playlist_id, u'Downloading page #%s' % page_num)
- try:
- response = json.loads(page)
- except ValueError as err:
- raise ExtractorError(u'Invalid JSON in API response: ' + compat_str(err))
+ more = self._download_json(
+ 'https://youtube.com/%s' % mobj.group('more'), playlist_id,
+ 'Downloading page #%s' % page_num,
+ transform_source=uppercase_escape)
+ content_html = more['content_html']
+ more_widget_html = more['load_more_widget_html']
- 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
- break
+ playlist_title = self._html_search_regex(
+ r'(?s)<h1 class="pl-header-title[^"]*">\s*(.*?)\s*</h1>',
+ page, u'title')
+
+ url_results = self._ids_to_results(ids)
+ return self.playlist_result(url_results, playlist_id, playlist_title)
- for entry in response['feed']['entry']:
- index = entry['yt$position']['$t']
- if 'media$group' in entry and 'media$player' in entry['media$group']:
- videos.append((index, entry['media$group']['media$player']['url']))
- videos = [v[1] for v in sorted(videos)]
+class YoutubeTopListIE(YoutubePlaylistIE):
+ IE_NAME = u'youtube:toplist'
+ IE_DESC = (u'YouTube.com top lists, "yttoplist:{channel}:{list title}"'
+ u' (Example: "yttoplist:music:Top Tracks")')
+ _VALID_URL = r'yttoplist:(?P<chann>.*?):(?P<title>.*?)$'
- url_results = [self.url_result(vurl, 'Youtube') for vurl in videos]
- return [self.playlist_result(url_results, playlist_id, playlist_title)]
+ def _real_extract(self, url):
+ mobj = re.match(self._VALID_URL, url)
+ channel = mobj.group('chann')
+ title = mobj.group('title')
+ query = compat_urllib_parse.urlencode({'title': title})
+ playlist_re = 'href="([^"]+?%s.*?)"' % re.escape(query)
+ channel_page = self._download_webpage('https://www.youtube.com/%s' % channel, title)
+ link = self._html_search_regex(playlist_re, channel_page, u'list')
+ url = compat_urlparse.urljoin('https://www.youtube.com/', link)
+
+ video_re = r'data-index="\d+".*?data-video-id="([0-9A-Za-z_-]{11})"'
+ ids = []
+ # sometimes the webpage doesn't contain the videos
+ # retry until we get them
+ for i in itertools.count(0):
+ msg = u'Downloading Youtube mix'
+ if i > 0:
+ msg += ', retry #%d' % i
+ webpage = self._download_webpage(url, title, msg)
+ ids = orderedSet(re.findall(video_re, webpage))
+ if ids:
+ break
+ url_results = self._ids_to_results(ids)
+ return self.playlist_result(url_results, playlist_title=title)
class YoutubeChannelIE(InfoExtractor):
IE_DESC = u'YouTube.com channels'
_VALID_URL = r"^(?:https?://)?(?:youtu\.be|(?:\w+\.)?youtube(?:-nocookie)?\.com)/channel/([0-9A-Za-z_-]+)"
- _TEMPLATE_URL = 'http://www.youtube.com/channel/%s/videos?sort=da&flow=list&view=0&page=%s&gl=US&hl=en'
_MORE_PAGES_INDICATOR = 'yt-uix-load-more'
- _MORE_PAGES_URL = 'http://www.youtube.com/c4_browse_ajax?action_load_more_videos=1&flow=list&paging=%s&view=0&sort=da&channel_id=%s'
+ _MORE_PAGES_URL = 'https://www.youtube.com/c4_browse_ajax?action_load_more_videos=1&flow=list&paging=%s&view=0&sort=da&channel_id=%s'
IE_NAME = u'youtube:channel'
def extract_videos_from_page(self, page):
# Download channel page
channel_id = mobj.group(1)
video_ids = []
- pagenum = 1
-
- url = self._TEMPLATE_URL % (channel_id, pagenum)
- page = self._download_webpage(url, channel_id,
- u'Downloading page #%s' % pagenum)
-
- # Extract video identifiers
- ids_in_page = self.extract_videos_from_page(page)
- video_ids.extend(ids_in_page)
-
- # Download any subsequent channel pages using the json-based channel_ajax query
- if self._MORE_PAGES_INDICATOR in page:
+ url = 'https://www.youtube.com/channel/%s/videos' % channel_id
+ channel_page = self._download_webpage(url, channel_id)
+ autogenerated = re.search(r'''(?x)
+ class="[^"]*?(?:
+ channel-header-autogenerated-label|
+ yt-channel-title-autogenerated
+ )[^"]*"''', channel_page) is not None
+
+ if autogenerated:
+ # The videos are contained in a single page
+ # the ajax pages can't be used, they are empty
+ video_ids = self.extract_videos_from_page(channel_page)
+ else:
+ # Download all channel pages using the json-based channel_ajax query
for pagenum in itertools.count(1):
url = self._MORE_PAGES_URL % (pagenum, channel_id)
- page = self._download_webpage(url, channel_id,
- u'Downloading page #%s' % pagenum)
-
- page = json.loads(page)
+ page = self._download_json(
+ url, channel_id, note=u'Downloading page #%s' % pagenum,
+ transform_source=uppercase_escape)
ids_in_page = self.extract_videos_from_page(page['content_html'])
video_ids.extend(ids_in_page)
-
- if self._MORE_PAGES_INDICATOR not in page['load_more_widget_html']:
+
+ if self._MORE_PAGES_INDICATOR not in page['load_more_widget_html']:
break
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):
IE_DESC = u'YouTube.com user videos (URL or "ytuser" keyword)'
- _VALID_URL = r'(?:(?:(?:https?://)?(?:\w+\.)?youtube\.com/user/)|ytuser:)([A-Za-z0-9_-]+)'
- _TEMPLATE_URL = 'http://gdata.youtube.com/feeds/api/users/%s'
+ _VALID_URL = r'(?:(?:(?:https?://)?(?:\w+\.)?youtube\.com/(?:user/)?(?!(?:attribution_link|watch|results)(?:$|[^a-z_A-Z0-9-])))|ytuser:)(?!feed/)([A-Za-z0-9_-]+)'
+ _TEMPLATE_URL = 'https://gdata.youtube.com/feeds/api/users/%s'
_GDATA_PAGE_SIZE = 50
- _GDATA_URL = 'http://gdata.youtube.com/feeds/api/users/%s/uploads?max-results=%d&start-index=%d'
- _VIDEO_INDICATOR = r'/watch\?v=(.+?)[\<&]'
+ _GDATA_URL = 'https://gdata.youtube.com/feeds/api/users/%s/uploads?max-results=%d&start-index=%d&alt=json'
IE_NAME = u'youtube:user'
+ @classmethod
+ def suitable(cls, url):
+ # Don't return True if the url can be extracted with other youtube
+ # extractor, the regex would is too permissive and it would match.
+ other_ies = iter(klass for (name, klass) in globals().items() if name.endswith('IE') and klass is not cls)
+ if any(ie.suitable(url) for ie in other_ies): return False
+ else: return super(YoutubeUserIE, cls).suitable(url)
+
def _real_extract(self, url):
# Extract username
mobj = re.match(self._VALID_URL, url)
# page by page until there are no video ids - it means we got
# all of them.
- video_ids = []
-
- for pagenum in itertools.count(0):
+ def download_page(pagenum):
start_index = pagenum * self._GDATA_PAGE_SIZE + 1
gdata_url = self._GDATA_URL % (username, self._GDATA_PAGE_SIZE, start_index)
- page = self._download_webpage(gdata_url, username,
- u'Downloading video ids from %d to %d' % (start_index, start_index + self._GDATA_PAGE_SIZE))
-
- # Extract video identifiers
- ids_in_page = []
+ page = self._download_webpage(
+ gdata_url, username,
+ u'Downloading video ids from %d to %d' % (
+ start_index, start_index + self._GDATA_PAGE_SIZE))
- for mobj in re.finditer(self._VIDEO_INDICATOR, page):
- if mobj.group(1) not in ids_in_page:
- ids_in_page.append(mobj.group(1))
-
- video_ids.extend(ids_in_page)
+ try:
+ response = json.loads(page)
+ except ValueError as err:
+ raise ExtractorError(u'Invalid JSON in API response: ' + compat_str(err))
+ if 'entry' not in response['feed']:
+ return
- # A little optimization - if current page is not
- # "full", ie. does not contain PAGE_SIZE video ids then
- # we can assume that this page is the last one - there
- # are no more ids on further pages - no need to query
- # again.
+ # Extract video identifiers
+ entries = response['feed']['entry']
+ for entry in entries:
+ title = entry['title']['$t']
+ video_id = entry['id']['$t'].split('/')[-1]
+ yield {
+ '_type': 'url',
+ 'url': video_id,
+ 'ie_key': 'Youtube',
+ 'id': video_id,
+ 'title': title,
+ }
+ url_results = PagedList(download_page, self._GDATA_PAGE_SIZE)
- if len(ids_in_page) < self._GDATA_PAGE_SIZE:
- break
+ return self.playlist_result(url_results, playlist_title=username)
- 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)]
class YoutubeSearchIE(SearchInfoExtractor):
IE_DESC = u'YouTube.com searches'
- _API_URL = 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc'
+ _API_URL = u'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc'
_MAX_RESULTS = 1000
IE_NAME = u'youtube:search'
_SEARCH_KEY = 'ytsearch'
- def report_download_page(self, query, pagenum):
- """Report attempt to download search page with given number."""
- self._downloader.to_screen(u'[youtube] query "%s": Downloading page %s' % (query, pagenum))
-
def _get_n_results(self, query, n):
"""Get a specified number of results for a query"""
video_ids = []
pagenum = 0
limit = n
-
- while (50 * pagenum) < limit:
- self.report_download_page(query, pagenum+1)
- result_url = self._API_URL % (compat_urllib_parse.quote_plus(query), (50*pagenum)+1)
- request = compat_urllib_request.Request(result_url)
- try:
- data = compat_urllib_request.urlopen(request).read().decode('utf-8')
- except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
- raise ExtractorError(u'Unable to download API page: %s' % compat_str(err))
- api_response = json.loads(data)['data']
-
- if not 'items' in api_response:
- raise ExtractorError(u'[youtube] No video results')
+ PAGE_SIZE = 50
+
+ while (PAGE_SIZE * pagenum) < limit:
+ result_url = self._API_URL % (
+ compat_urllib_parse.quote_plus(query.encode('utf-8')),
+ (PAGE_SIZE * pagenum) + 1)
+ data_json = self._download_webpage(
+ result_url, video_id=u'query "%s"' % query,
+ note=u'Downloading page %s' % (pagenum + 1),
+ errnote=u'Unable to download API page')
+ data = json.loads(data_json)
+ api_response = data['data']
+
+ if 'items' not in api_response:
+ raise ExtractorError(
+ u'[youtube] No video results', expected=True)
new_ids = list(video['id'] for video in api_response['items'])
video_ids += new_ids
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'
+
+
+class YoutubeSearchURLIE(InfoExtractor):
+ IE_DESC = u'YouTube.com search URLs'
+ IE_NAME = u'youtube:search_url'
+ _VALID_URL = r'https?://(?:www\.)?youtube\.com/results\?(.*?&)?search_query=(?P<query>[^&]+)(?:[&]|$)'
+
+ def _real_extract(self, url):
+ mobj = re.match(self._VALID_URL, url)
+ query = compat_urllib_parse.unquote_plus(mobj.group('query'))
+
+ webpage = self._download_webpage(url, query)
+ result_code = self._search_regex(
+ r'(?s)<ol class="item-section"(.*?)</ol>', webpage, u'result HTML')
+
+ part_codes = re.findall(
+ r'(?s)<h3 class="yt-lockup-title">(.*?)</h3>', result_code)
+ entries = []
+ for part_code in part_codes:
+ part_title = self._html_search_regex(
+ [r'(?s)title="([^"]+)"', r'>([^<]+)</a>'], part_code, 'item title', fatal=False)
+ part_url_snippet = self._html_search_regex(
+ r'(?s)href="([^"]+)"', part_code, 'item URL')
+ part_url = compat_urlparse.urljoin(
+ 'https://www.youtube.com/', part_url_snippet)
+ entries.append({
+ '_type': 'url',
+ 'url': part_url,
+ 'title': part_title,
+ })
+
+ return {
+ '_type': 'playlist',
+ 'entries': entries,
+ 'title': query,
+ }
+
+
class YoutubeShowIE(InfoExtractor):
IE_DESC = u'YouTube.com (multi-season) shows'
_VALID_URL = r'https?://www\.youtube\.com/show/(.*)'
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
action = 'action_load_system_feed'
if self._PERSONAL_FEED:
action = 'action_load_personal_feed'
- return 'http://www.youtube.com/feed_ajax?%s=1&feed_name=%s&paging=%%s' % (action, self._FEED_NAME)
+ return 'https://www.youtube.com/feed_ajax?%s=1&feed_name=%s&paging=%%s' % (action, self._FEED_NAME)
@property
def IE_NAME(self):
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
- info = self._download_webpage(self._FEED_TEMPLATE % paging,
+ paging = 0
+ for i in itertools.count(1):
+ info = self._download_json(self._FEED_TEMPLATE % paging,
u'%s feed' % self._FEED_NAME,
u'Downloading page %s' % i)
- info = json.loads(info)
- feed_html = info['feed_html']
+ feed_html = info.get('feed_html') or info.get('content_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)
- if info['paging'] is None:
+ feed_entries.extend(
+ self.url_result(video_id, 'Youtube', video_id=video_id)
+ for video_id in ids)
+ mobj = re.search(
+ r'data-uix-load-more-href="/?[^"]+paging=(?P<paging>\d+)',
+ feed_html)
+ if mobj is None:
break
+ paging = mobj.group('paging')
return self.playlist_result(feed_entries, playlist_title=self._PLAYLIST_TITLE)
class YoutubeSubscriptionsIE(YoutubeFeedsInfoExtractor):
- IE_DESC = u'YouTube.com subscriptions feed, "ytsubs" keyword(requires authentication)'
+ IE_DESC = u'YouTube.com subscriptions feed, "ytsubs" keyword (requires authentication)'
_VALID_URL = r'https?://www\.youtube\.com/feed/subscriptions|:ytsubs(?:criptions)?'
_FEED_NAME = 'subscriptions'
_PLAYLIST_TITLE = u'Youtube Subscriptions'
_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)'
- _VALID_URL = r'https?://www\.youtube\.com/my_favorites|:ytfav(?:o?rites)?'
+ _VALID_URL = r'https?://www\.youtube\.com/my_favorites|:ytfav(?:ou?rites)?'
_LOGIN_REQUIRED = True
def _real_extract(self, url):
webpage = self._download_webpage('https://www.youtube.com/my_favorites', 'Youtube Favourites videos')
playlist_id = self._search_regex(r'list=(.+?)["&]', webpage, u'favourites playlist id')
return self.url_result(playlist_id, 'YoutubePlaylist')
+
+
+class YoutubeTruncatedURLIE(InfoExtractor):
+ IE_NAME = 'youtube:truncated_url'
+ IE_DESC = False # Do not list
+ _VALID_URL = r'''(?x)
+ (?:https?://)?[^/]+/watch\?(?:
+ feature=[a-z_]+|
+ annotation_id=annotation_[^&]+
+ )?$|
+ (?:https?://)?(?:www\.)?youtube\.com/attribution_link\?a=[^&]+$
+ '''
+
+ _TESTS = [{
+ 'url': 'http://www.youtube.com/watch?annotation_id=annotation_3951667041',
+ 'only_matching': True,
+ }, {
+ 'url': 'http://www.youtube.com/watch?',
+ 'only_matching': True,
+ }]
+
+ def _real_extract(self, url):
+ raise ExtractorError(
+ u'Did you forget to quote the URL? Remember that & is a meta '
+ u'character in most shells, so you want to put the URL in quotes, '
+ u'like youtube-dl '
+ u'"http://www.youtube.com/watch?feature=foo&v=BaW_jenozKc" '
+ u' or simply youtube-dl BaW_jenozKc .',
+ expected=True)
Rapsys Git / youtubedl / blobdiff