From 98951bbe76e63c8e5223e25aeefed03aa0df4ab7 Mon Sep 17 00:00:00 2001 From: unknown <7951720+u-n-k-n-o-w-n@users.noreply.github.com> Date: Sat, 20 Aug 2022 11:27:13 +0900 Subject: [PATCH] [options] Added workaround option to execute "n_function" --- test/test_youtube_signature.py | 14 +++++- youtube_dl/__init__.py | 1 + youtube_dl/compat.py | 6 +++ youtube_dl/extractor/youtube.py | 75 +++++++++++++++++++++++++++++++++ youtube_dl/options.py | 3 ++ 5 files changed, 98 insertions(+), 1 deletion(-) diff --git a/test/test_youtube_signature.py b/test/test_youtube_signature.py index cc18d0f7b..f9e520a0b 100644 --- a/test/test_youtube_signature.py +++ b/test/test_youtube_signature.py @@ -214,7 +214,10 @@ def t_factory(name, sig_func, url_pattern): test_id = m.group('id') def test_func(self): - basename = 'player-{0}-{1}.js'.format(name, test_id) + tn = name + if name.endswith('_wd'): + tn = name[:-3] + basename = 'player-{0}-{1}.js'.format(tn, test_id) fn = os.path.join(self.TESTDATA_DIR, basename) if not os.path.exists(fn): @@ -241,6 +244,10 @@ def n_sig(jscode, sig_input): return JSInterpreter(jscode).call_function(funcname, sig_input) +def n_sig_wd(jscode, sig_input): + return YoutubeIE(FakeYDL())._call_n_function_with_webdriver('chrome', jscode, sig_input) + + make_sig_test = t_factory( 'signature', signature, re.compile(r'.*-(?P[a-zA-Z0-9_-]+)(?:/watch_as3|/html5player)?\.[a-z]+$')) for test_spec in _SIG_TESTS: @@ -251,6 +258,11 @@ make_nsig_test = t_factory( for test_spec in _NSIG_TESTS: make_nsig_test(*test_spec) +make_nsig_wd_test = t_factory( + 'nsig_wd', n_sig_wd, re.compile(r'.+/player/(?P[a-zA-Z0-9_-]+)/.+.js$')) +for test_spec in _NSIG_TESTS: + make_nsig_wd_test(*test_spec) + if __name__ == '__main__': unittest.main() diff --git a/youtube_dl/__init__.py b/youtube_dl/__init__.py index 06bdfb689..b11faefef 100644 --- a/youtube_dl/__init__.py +++ b/youtube_dl/__init__.py @@ -419,6 +419,7 @@ def _real_main(argv=None): 'call_home': opts.call_home, 'sleep_interval': opts.sleep_interval, 'max_sleep_interval': opts.max_sleep_interval, + 'webdriver': opts.webdriver, 'external_downloader': opts.external_downloader, 'list_thumbnails': opts.list_thumbnails, 'playlist_items': opts.playlist_items, diff --git a/youtube_dl/compat.py b/youtube_dl/compat.py index ed1a33cf2..b7c5c3800 100644 --- a/youtube_dl/compat.py +++ b/youtube_dl/compat.py @@ -2448,6 +2448,11 @@ try: except ImportError: import BaseHTTPServer as compat_http_server +try: + from urllib.parse import quote as compat_urllib_quote +except ImportError: # Python 2 + from urllib import quote as compat_urllib_quote + try: from urllib.parse import unquote_to_bytes as compat_urllib_parse_unquote_to_bytes from urllib.parse import unquote as compat_urllib_parse_unquote @@ -3560,6 +3565,7 @@ __all__ = [ 'compat_tokenize_tokenize', 'compat_urllib_error', 'compat_urllib_parse', + 'compat_urllib_quote', 'compat_urllib_request', 'compat_urllib_request_DataHandler', 'compat_urllib_response', diff --git a/youtube_dl/extractor/youtube.py b/youtube_dl/extractor/youtube.py index 2e31a8979..5c5b33010 100644 --- a/youtube_dl/extractor/youtube.py +++ b/youtube_dl/extractor/youtube.py @@ -9,6 +9,7 @@ import os.path import random import re import traceback +import importlib from .common import InfoExtractor, SearchInfoExtractor from ..compat import ( @@ -19,6 +20,7 @@ from ..compat import ( compat_urllib_parse, compat_urllib_parse_parse_qs as compat_parse_qs, compat_urllib_parse_unquote_plus, + compat_urllib_quote, compat_urllib_parse_urlparse, compat_zip as zip, ) @@ -1464,6 +1466,11 @@ class YoutubeIE(YoutubeBaseInfoExtractor): super(YoutubeIE, self).__init__(*args, **kwargs) self._code_cache = {} self._player_cache = {} + self._webdriver = None + + def __del__(self): + if self._webdriver is not None: + self._webdriver.quit() # *ytcfgs, webpage=None def _extract_player_url(self, *ytcfgs, **kw_webpage): @@ -1633,6 +1640,22 @@ class YoutubeIE(YoutubeBaseInfoExtractor): if player_url is None: raise ExtractorError('Cannot decrypt nsig without player_url') + webdriver_type = self._downloader.params.get('webdriver', None) + if webdriver_type is not None: + try: + jscode = self._load_player(video_id, player_url) + ret = self._call_n_function_with_webdriver(webdriver_type, jscode, n) + except Exception as e: + self.report_warning( + '%s (%s %s)' % ( + 'Unable to decode n-parameter: download likely to be throttled', + error_to_compat_str(e), + traceback.format_exc()), + video_id=video_id) + return + self.write_debug('Decrypted nsig(with webdriver) {0} => {1}'.format(n, ret)) + return ret + try: jsi, player_id, func_code = self._extract_n_function_code(video_id, player_url) except ExtractorError as e: @@ -1656,6 +1679,58 @@ class YoutubeIE(YoutubeBaseInfoExtractor): self.write_debug('Decrypted nsig {0} => {1}'.format(n, ret)) return ret + def _call_n_function_with_webdriver(self, webdriver_type, jscode, n_param): + if self._webdriver is None: + wd = importlib.import_module('selenium.webdriver') + if webdriver_type == 'firefox': # geckodriver + o = wd.FirefoxOptions() + o.headless = True + s = wd.firefox.service.Service(log_path=os.path.devnull) + self._webdriver = wd.Firefox(options=o, service=s) + elif webdriver_type == 'chrome': # chromedriver + o = wd.ChromeOptions() + o.headless = True + """ + If you are using the snap version of the chromium, chromedriver is included in the snap package. + You should use that driver. + $ cd /snap/bin && sudo ln -s -T chromium.chromedriver chromedriver + or + s = wd.chrome.service.Service(executable_path='chromium.chromedriver') + self._webdriver = wd.Chrome(options=o, service=s) + """ + self._webdriver = wd.Chrome(options=o) + elif webdriver_type == 'edge': # msedgedriver + o = wd.EdgeOptions() + o.headless = True + self._webdriver = wd.Edge(options=o) + elif webdriver_type == 'safari': # safaridriver + """ + safaridriver does not have headless-mode. :( + But macOS includes safaridriver by default. + To enable automation on safaridriver, run the following command once from the admin terminal. + # safaridriver --enable + """ + self._webdriver = wd.Safari() + else: + raise ExtractorError('unsupported webdriver type: %s' % (webdriver_type)) + self._webdriver.get('about:blank') + funcname = self._extract_n_function_name(jscode) + alphabet = 'abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ' + dummyfunc = ''.join(random.choice(alphabet) for _ in range(8)) + f = ('return ((e) => {{' + 'const d = decodeURIComponent(e);' + 'const p = d.lastIndexOf("}}");' + 'const th = d.substring(0, p);' + 'const bh = d.substring(p);' + 'const m = "var {0};" + th + ";{0} = {1};" + bh;' + 'const s = document.createElement("script");' + 's.innerHTML = m;' + 'document.body.append(s);' + 'return {0}("{2}");' + '}})("{3}");').format(dummyfunc, funcname, n_param, compat_urllib_quote(jscode)) + n = self._webdriver.execute_script(f) + return n + def _extract_n_function_name(self, jscode): func_name, idx = self._search_regex( # new: (b=String.fromCharCode(110),c=a.get(b))&&c=nfunc[idx](c) diff --git a/youtube_dl/options.py b/youtube_dl/options.py index 61705d1f0..27dcc4807 100644 --- a/youtube_dl/options.py +++ b/youtube_dl/options.py @@ -576,6 +576,9 @@ def parseOpts(overrideArguments=None): 'Upper bound of a range for randomized sleep before each download ' '(maximum possible number of seconds to sleep). Must only be used ' 'along with --min-sleep-interval.')) + workarounds.add_option( + '--webdriver', metavar='TYPE', dest='webdriver', default=None, + help='Specify webdriver type when you want to use selenium webdriver to execute "n_function" : "firefox", "chrome", "edge", or "safari"') verbosity = optparse.OptionGroup(parser, 'Verbosity / Simulation Options') verbosity.add_option(