X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/71050031bb901b19a829cc44763bf68e67b1f648..c9774cf13273185f8f8075590dce493a18ead418:/youtube_dl/extractor/youtube.py diff --git a/youtube_dl/extractor/youtube.py b/youtube_dl/extractor/youtube.py index c992cba..7c50881 100644 --- a/youtube_dl/extractor/youtube.py +++ b/youtube_dl/extractor/youtube.py @@ -7,20 +7,16 @@ import itertools import json import os.path import re -import socket -import string import struct import traceback -import xml.etree.ElementTree import zlib from .common import InfoExtractor, SearchInfoExtractor from .subtitles import SubtitlesInfoExtractor +from ..jsinterp import JSInterpreter from ..utils import ( compat_chr, - compat_http_client, compat_parse_qs, - compat_urllib_error, compat_urllib_parse, compat_urllib_request, compat_urlparse, @@ -29,35 +25,31 @@ from ..utils import ( 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() @@ -67,12 +59,12 @@ class YoutubeBaseInfoExtractor(InfoExtractor): 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 = self._search_regex(r'(?s)]* 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)]* 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): @@ -139,191 +131,106 @@ class YoutubeBaseInfoExtractor(InfoExtractor): class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): IE_DESC = u'YouTube.com' - _VALID_URL = r"""^ + _VALID_URL = r"""(?x)^ ( - (?:https?://)? # http(s):// (optional) - (?:(?:(?:(?:\w+\.)?youtube(?:-nocookie)?\.com/| + (?: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= ) )) |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_-]{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', '36', '17', '13', - # Apple HTTP Live Streaming - '96', '95', '94', '93', '92', '132', '151', - # 3D - '85', '84', '102', '83', '101', '82', '100', - # Dash video - '138', '137', '248', '136', '247', '135', '246', - '245', '244', '134', '243', '133', '242', '160', - # Dash audio - '141', '172', '140', '171', '139', - ] - _available_formats_prefer_free = ['38', '46', '37', '45', '22', '44', '35', '43', '34', '18', '6', '5', '36', '17', '13', - # Apple HTTP Live Streaming - '96', '95', '94', '93', '92', '132', '151', - # 3D - '85', '102', '84', '101', '83', '100', '82', - # Dash video - '138', '248', '137', '247', '136', '246', '245', - '244', '135', '243', '134', '242', '133', '160', - # Dash audio - '172', '141', '171', '140', '139', - ] - _video_formats_map = { - 'flv': ['35', '34', '6', '5'], - '3gp': ['36', '17', '13'], - 'mp4': ['38', '37', '22', '18'], - 'webm': ['46', '45', '44', '43'], - } - _video_extensions = { - '13': '3gp', - '17': '3gp', - '18': 'mp4', - '22': 'mp4', - '36': '3gp', - '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', + '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': 'mp4', - '93': 'mp4', - '94': 'mp4', - '95': 'mp4', - '96': 'mp4', - '132': 'mp4', - '151': 'mp4', - - # Dash mp4 - '133': 'mp4', - '134': 'mp4', - '135': 'mp4', - '136': 'mp4', - '137': 'mp4', - '138': 'mp4', - '160': 'mp4', + '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': 'm4a', - '140': 'm4a', - '141': 'm4a', + '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 - '171': 'webm', - '172': 'webm', - '242': 'webm', - '243': 'webm', - '244': 'webm', - '245': 'webm', - '246': 'webm', - '247': 'webm', - '248': 'webm', - } - _video_dimensions = { - '5': '240x400', - '6': '???', - '13': '???', - '17': '144x176', - '18': '360x640', - '22': '720x1280', - '34': '360x640', - '35': '480x854', - '36': '240x320', - '37': '1080x1920', - '38': '3072x4096', - '43': '360x640', - '44': '480x854', - '45': '720x1280', - '46': '1080x1920', - '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', - '133': '240p', - '134': '360p', - '135': '480p', - '136': '720p', - '137': '1080p', - '138': '>1080p', - '139': '48k', - '140': '128k', - '141': '256k', - '160': '192p', - '171': '128k', - '172': '256k', - '242': '240p', - '243': '360p', - '244': '480p', - '245': '480p', - '246': '480p', - '247': '720p', - '248': '1080p', - } - _special_itags = { - '82': '3D', - '83': '3D', - '84': '3D', - '85': '3D', - '100': '3D', - '101': '3D', - '102': '3D', - '133': 'DASH Video', - '134': 'DASH Video', - '135': 'DASH Video', - '136': 'DASH Video', - '137': 'DASH Video', - '138': 'DASH Video', - '139': 'DASH Audio', - '140': 'DASH Audio', - '141': 'DASH Audio', - '160': 'DASH Video', - '171': 'DASH Audio', - '172': 'DASH Audio', - '242': 'DASH Video', - '243': 'DASH Video', - '244': 'DASH Video', - '245': 'DASH Video', - '246': 'DASH Video', - '247': 'DASH Video', - '248': 'DASH Video', + '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}, + + # 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'}, } IE_NAME = u'youtube' @@ -336,7 +243,8 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): u"uploader": u"Philipp Hagemeister", u"uploader_id": u"phihag", u"upload_date": u"20121002", - u"description": u"test chars: \"'/\\ä↭𝕐\n\nThis is a test video for youtube-dl.\n\nFor more information, contact phihag@phihag.de ." + u"description": u"test chars: \"'/\\ä↭𝕐\ntest URL: https://github.com/rg3/youtube-dl/issues/1892\n\nThis is a test video for youtube-dl.\n\nFor more information, contact phihag@phihag.de .", + u"categories": [u'Science & Technology'], } }, { @@ -346,7 +254,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): u"info_dict": { u"upload_date": u"20120506", u"title": u"Icona Pop - I Love It (feat. Charli XCX) [OFFICIAL VIDEO]", - u"description": u"md5:5b292926389560516e384ac437c0ec07", + u"description": u"md5:fea86fda2d5a5784273df5c7cc994d9f", u"uploader": u"Icona Pop", u"uploader_id": u"IconaPop" } @@ -363,6 +271,51 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): u"uploader_id": u"justintimberlakeVEVO" } }, + { + 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'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'youtube_include_dash_manifest': True, + u'format': '141', + }, + }, ] @@ -370,16 +323,12 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): def suitable(cls, url): """Receives a URL and returns True if suitable for this IE.""" if YoutubePlaylistIE.suitable(url): return False - return re.match(cls._VALID_URL, url, re.VERBOSE) is not None + return re.match(cls._VALID_URL, url) is not None def __init__(self, *args, **kwargs): super(YoutubeIE, self).__init__(*args, **kwargs) self._player_cache = {} - 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 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) @@ -492,113 +441,10 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): def _parse_sig_js(self, jscode): funcname = self._search_regex( r'signature=([a-zA-Z]+)', jscode, - u'Initial JS player signature function name') - - functions = {} - - def argidx(varname): - return string.lowercase.index(varname) - - def interpret_statement(stmt, local_vars, allow_recursion=20): - if allow_recursion < 0: - raise ExtractorError(u'Recursion limit reached') - - if stmt.startswith(u'var '): - stmt = stmt[len(u'var '):] - ass_m = re.match(r'^(?P[a-z]+)(?:\[(?P[^\]]+)\])?' + - r'=(?P.*)$', stmt) - if ass_m: - if ass_m.groupdict().get('index'): - def assign(val): - lvar = local_vars[ass_m.group('out')] - idx = interpret_expression(ass_m.group('index'), - local_vars, allow_recursion) - assert isinstance(idx, int) - lvar[idx] = val - return val - expr = ass_m.group('expr') - else: - def assign(val): - local_vars[ass_m.group('out')] = val - return val - expr = ass_m.group('expr') - elif stmt.startswith(u'return '): - assign = lambda v: v - expr = stmt[len(u'return '):] - else: - raise ExtractorError( - u'Cannot determine left side of statement in %r' % stmt) - - v = interpret_expression(expr, local_vars, allow_recursion) - return assign(v) - - def interpret_expression(expr, local_vars, allow_recursion): - if expr.isdigit(): - return int(expr) - - if expr.isalpha(): - return local_vars[expr] - - m = re.match(r'^(?P[a-z]+)\.(?P.*)$', expr) - if m: - member = m.group('member') - val = local_vars[m.group('in')] - if member == 'split("")': - return list(val) - if member == 'join("")': - return u''.join(val) - if member == 'length': - return len(val) - if member == 'reverse()': - return val[::-1] - slice_m = re.match(r'slice\((?P.*)\)', member) - if slice_m: - idx = interpret_expression( - slice_m.group('idx'), local_vars, allow_recursion-1) - return val[idx:] - - m = re.match( - r'^(?P[a-z]+)\[(?P.+)\]$', expr) - if m: - val = local_vars[m.group('in')] - idx = interpret_expression(m.group('idx'), local_vars, - allow_recursion-1) - return val[idx] - - m = re.match(r'^(?P.+?)(?P[%])(?P.+?)$', expr) - if m: - a = interpret_expression(m.group('a'), - local_vars, allow_recursion) - b = interpret_expression(m.group('b'), - local_vars, allow_recursion) - return a % b - - m = re.match( - r'^(?P[a-zA-Z]+)\((?P[a-z0-9,]+)\)$', expr) - if m: - fname = m.group('func') - if fname not in functions: - functions[fname] = extract_function(fname) - argvals = [int(v) if v.isdigit() else local_vars[v] - for v in m.group('args').split(',')] - return functions[fname](argvals) - raise ExtractorError(u'Unsupported JS expression %r' % expr) - - def extract_function(funcname): - func_m = re.search( - r'function ' + re.escape(funcname) + - r'\((?P[a-z,]+)\){(?P[^}]+)}', - jscode) - argnames = func_m.group('args').split(',') - - def resf(args): - local_vars = dict(zip(argnames, args)) - for stmt in func_m.group('code').split(';'): - res = interpret_statement(stmt, local_vars) - return res - return resf - - initial_function = extract_function(funcname) + 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): @@ -1019,6 +865,8 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): """Turn the encrypted s field into a working signature""" if player_url is not None: + if player_url.startswith(u'//'): + player_url = u'https:' + player_url try: player_id = (player_url, len(s)) if player_id not in self._player_cache: @@ -1085,7 +933,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): def _get_available_subtitles(self, video_id, webpage): try: sub_list = self._download_webpage( - 'http://video.google.com/timedtext?hl=en&type=list&v=%s' % video_id, + '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)) @@ -1098,10 +946,10 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): params = compat_urllib_parse.urlencode({ 'lang': lang, 'v': video_id, - 'fmt': self._downloader.params.get('subtitlesformat'), - 'name': l[0].encode('utf-8'), + 'fmt': self._downloader.params.get('subtitlesformat', 'srt'), + 'name': unescapeHTML(l[0]).encode('utf-8'), }) - url = u'http://www.youtube.com/api/timedtext?' + params + url = u'https://www.youtube.com/api/timedtext?' + params sub_lang_list[lang] = url if not sub_lang_list: self._downloader.report_warning(u'video doesn\'t have subtitles') @@ -1111,7 +959,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): 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_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' % video_id @@ -1130,8 +978,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): 'asrs': 1, }) list_url = caption_url + '&' + list_params - list_page = self._download_webpage(list_url, video_id) - caption_list = xml.etree.ElementTree.fromstring(list_page.encode('utf-8')) + caption_list = self._download_xml(list_url, video_id) original_lang_node = caption_list.find('track') if original_lang_node is None or original_lang_node.attrib.get('kind') != 'asr' : self._downloader.report_warning(u'Video doesn\'t have automatic captions') @@ -1156,69 +1003,14 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): self._downloader.report_warning(err_msg) return {} - 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, '???'), - ' ('+self._special_itags[x]+')' if x in self._special_itags else '')) - - def _extract_id(self, url): - mobj = re.match(self._VALID_URL, url, re.VERBOSE) + @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. - # Format can be specified as itag or 'mp4' or 'flv' etc. We pick the highest quality - # available in the specified format. For example, - # if '1/2/3/4' is requested and '2' and '4' are available, we pick '2'. - # if '1/mp4/3/4' is requested and '1' and '5' (is a mp4) are available, we pick '1'. - # if '1/mp4/3/4' is requested and '4' and '5' (is a mp4) are available, we pick '5'. - 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 rf in self._video_formats_map: - for srf in self._video_formats_map[rf]: - if srf in url_map: - video_url_list = [(srf, url_map[srf])] - break - else: - continue - 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): @@ -1238,22 +1030,19 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): return self._download_webpage(url, video_id, note=u'Searching for annotations.', errnote=u'Unable to download video annotations.') def _real_extract(self, url): + 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.*?"(https?:\\/\\/.*?watch.*?-.*?\.swf)"', video_webpage) @@ -1270,14 +1059,14 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): # 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') @@ -1285,7 +1074,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): 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, @@ -1295,9 +1084,18 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): 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: @@ -1321,7 +1119,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): # title if 'title' in video_info: - video_title = compat_urllib_parse.unquote_plus(video_info['title'][0]) + video_title = video_info['title'][0] else: self._downloader.report_warning(u'Unable to extract video title') video_title = u'_' @@ -1340,14 +1138,36 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): # upload date upload_date = None - mobj = re.search(r'id="eow-date.*?>(.*?)', video_webpage, re.DOTALL) + mobj = re.search(r'(?s)id="eow-date.*?>(.*?)', video_webpage) + if mobj is None: + mobj = re.search( + r'(?s)id="watch-uploader-info".*?>.*?(?:Published|Uploaded|Streamed live) on (.*?)', + 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)(.*?)', 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) + + [^<]+ + + ''', r'\1', video_description) video_description = clean_html(video_description) else: fd_mobj = re.search(r'([\d,]+)' % 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 = self.extract_subtitles(video_id, video_webpage) @@ -1365,9 +1195,9 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): 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])) # annotations video_annotations = None @@ -1375,13 +1205,13 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): 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 if 'url_encoded_fmt_stream_map' not in args: @@ -1400,9 +1230,27 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): 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])] + 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: @@ -1447,52 +1295,88 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): 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, hlsvp or url_encoded_fmt_stream_map information found in video info') - results = [] - for itag, video_real_url in video_url_list: - # Extension - video_extension = self._video_extensions.get(itag, 'flv') - - video_format = '{0} - {1}{2}'.format(itag if itag else video_extension, - self._video_dimensions.get(itag, '???'), - ' ('+self._special_itags[itag]+')' if itag in self._special_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, - 'format_id': itag, - 'thumbnail': video_thumbnail, - 'description': video_description, - 'player_url': player_url, - 'subtitles': video_subtitles, - 'duration': video_duration, - 'age_limit': 18 if age_gate else 0, - 'annotations': video_annotations, - 'webpage_url': 'https://www.youtube.com/watch?v=%s' % video_id, - }) - return results + # 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/ @@ -1501,23 +1385,46 @@ class YoutubePlaylistIE(InfoExtractor): \? (?:.*?&)*? (?:p|a|list)= | p/ ) - ((?:PL|EC|UU|FL)?[0-9A-Za-z-_]{10,}) + ( + (?:PL|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|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[0-9A-Za-z_-]{11})&[^"]*?index=(?P\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) playlist_id = mobj.group(1) or mobj.group(2) @@ -1528,52 +1435,94 @@ class YoutubePlaylistIE(InfoExtractor): video_id = query_dict['v'][0] if self._downloader.params.get('noplaylist'): self.to_screen(u'Downloading just video %s because of --no-playlist' % video_id) - return self.url_result('https://www.youtube.com/watch?v=' + video_id, 'Youtube') + return self.url_result(video_id, 'Youtube', video_id=video_id) else: - self.to_screen(u'Downloading playlist PL%s - add --no-playlist to just download video %s' % (playlist_id, video_id)) + 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) - # Download playlist videos from API - videos = [] + 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'
[^<]*?(The|This) playlist (does not exist|is private)[^<]*?
', 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_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)

\s*(.*?)\s*

', + page, u'title') - for entry in response['feed']['entry']: - index = entry['yt$position']['$t'] - if 'media$group' in entry and 'yt$videoid' in entry['media$group']: - videos.append(( - index, - 'https://www.youtube.com/watch?v=' + entry['media$group']['yt$videoid']['$t'] - )) + url_results = self._ids_to_results(ids) + return self.playlist_result(url_results, playlist_id, playlist_title) - videos = [v[1] for v in sorted(videos)] - url_results = [self.url_result(vurl, 'Youtube') for vurl in videos] - return [self.playlist_result(url_results, playlist_id, playlist_title)] +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.*?):(?P.*?)$' + + 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_-]+)" _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): @@ -1592,34 +1541,45 @@ class YoutubeChannelIE(InfoExtractor): # Download channel page channel_id = mobj.group(1) video_ids = [] - - # 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) - - 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']: - break + 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_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']: + 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/)?(?!(?:attribution_link|watch)(?:$|[^a-z_A-Z0-9-])))|ytuser:)(?!feed/)([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)(?:$|[^a-z_A-Z0-9-])))|ytuser:)(?!feed/)([A-Za-z0-9_-]+)' + _TEMPLATE_URL = 'https://gdata.youtube.com/feeds/api/users/%s' _GDATA_PAGE_SIZE = 50 - _GDATA_URL = 'http://gdata.youtube.com/feeds/api/users/%s/uploads?max-results=%d&start-index=%d&alt=json' + _GDATA_URL = 'https://gdata.youtube.com/feeds/api/users/%s/uploads?max-results=%d&start-index=%d&alt=json' IE_NAME = u'youtube:user' @classmethod @@ -1643,72 +1603,68 @@ class YoutubeUserIE(InfoExtractor): # page by page until there are no video ids - it means we got # all of them. - video_ids = [] - - for pagenum in itertools.count(0): + def download_page(pagenum): start_index = pagenum * self._GDATA_PAGE_SIZE + 1 gdata_url = self._GDATA_URL % (username, self._GDATA_PAGE_SIZE, start_index) - page = self._download_webpage(gdata_url, username, - u'Downloading video ids from %d to %d' % (start_index, start_index + self._GDATA_PAGE_SIZE)) + page = self._download_webpage( + gdata_url, username, + u'Downloading video ids from %d to %d' % ( + start_index, start_index + self._GDATA_PAGE_SIZE)) try: response = json.loads(page) except ValueError as err: raise ExtractorError(u'Invalid JSON in API response: ' + compat_str(err)) if 'entry' not in response['feed']: - # Number of videos is a multiple of self._MAX_RESULTS - break + return # Extract video identifiers - ids_in_page = [] - for entry in response['feed']['entry']: - ids_in_page.append(entry['id']['$t'].split('/')[-1]) - video_ids.extend(ids_in_page) - - # 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. - - if len(ids_in_page) < self._GDATA_PAGE_SIZE: - break + 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) + + 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 @@ -1718,14 +1674,54 @@ class YoutubeSearchIE(SearchInfoExtractor): 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 id="search-results"(.*?)</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="([^"]+)"', 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/(.*)' @@ -1748,7 +1744,6 @@ class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor): 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 @@ -1757,7 +1752,7 @@ class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor): 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): @@ -1768,23 +1763,27 @@ class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor): 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' @@ -1800,9 +1799,15 @@ class YoutubeWatchLaterIE(YoutubeFeedsInfoExtractor): _VALID_URL = r'https?://www\.youtube\.com/feed/watch_later|:ytwatchlater' _FEED_NAME = 'watch_later' _PLAYLIST_TITLE = u'Youtube Watch Later' - _PAGING_STEP = 100 _PERSONAL_FEED = True +class YoutubeHistoryIE(YoutubeFeedsInfoExtractor): + IE_DESC = u'Youtube watch history, "ythistory" keyword (requires authentication)' + _VALID_URL = u'https?://www\.youtube\.com/feed/history|:ythistory' + _FEED_NAME = 'history' + _PERSONAL_FEED = True + _PLAYLIST_TITLE = u'Youtube Watch History' + class YoutubeFavouritesIE(YoutubeBaseInfoExtractor): IE_NAME = u'youtube:favorites' IE_DESC = u'YouTube.com favourite videos, "ytfav" keyword (requires authentication)' @@ -1818,13 +1823,16 @@ class YoutubeFavouritesIE(YoutubeBaseInfoExtractor): class YoutubeTruncatedURLIE(InfoExtractor): IE_NAME = 'youtube:truncated_url' IE_DESC = False # Do not list - _VALID_URL = r'(?:https?://)?[^/]+/watch\?feature=[a-z_]+$' + _VALID_URL = r'''(?x) + (?:https?://)?[^/]+/watch\?(?:feature=[a-z_]+)?$| + (?:https?://)?(?:www\.)?youtube\.com/attribution_link\?a=[^&]+$ + ''' 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 ).', + u'"http://www.youtube.com/watch?feature=foo&v=BaW_jenozKc" ' + u' or simply youtube-dl BaW_jenozKc .', expected=True)