diff --git a/test/test_jsinterp_external.py b/test/test_jsi_external.py similarity index 77% rename from test/test_jsinterp_external.py rename to test/test_jsi_external.py index 081f4257af..1d52e3fb39 100644 --- a/test/test_jsinterp_external.py +++ b/test/test_jsi_external.py @@ -55,19 +55,36 @@ class TestExternalJSI(unittest.TestCase): maxDiff = 2000 def setUp(self): + print() self.ydl = FakeYDL() - self.url = '' + self.url_param = '' if not self._JSI_CLASS.exe_version: print(f'{self._JSI_CLASS.__name__} is not installed, skipping') self.skipTest('Not available') @property def jsi(self): - return self._JSI_CLASS(self.ydl, self.url, 10, {}) + return self._JSI_CLASS(self.ydl, self.url_param, 10, {}) def test_execute(self): self.assertEqual(self.jsi.execute('console.log("Hello, world!");'), 'Hello, world!') + def test_user_agent(self): + ua = self.ydl.params['http_headers']['User-Agent'] + self.assertEqual(self.jsi.execute('console.log(navigator.userAgent);'), ua) + self.assertNotEqual(self.jsi.execute('console.log(JSON.stringify(navigator.webdriver));'), 'true') + + jsi = self._JSI_CLASS(self.ydl, self.url_param, 10, {}, user_agent='test/ua') + self.assertEqual(jsi.execute('console.log(navigator.userAgent);'), 'test/ua') + + def test_location(self): + if 'location' not in self._JSI_CLASS._SUPPORTED_FEATURES: + print(f'{self._JSI_CLASS.__name__} does not support location, skipping') + self.skipTest('Location not supported') + self.url_param = 'https://example.com/123/456' + self.assertEqual(self.jsi.execute('console.log(JSON.stringify([location.href, location.hostname]));'), + '["https://example.com/123/456","example.com"]') + def test_execute_dom_parse(self): if 'dom' not in self.jsi._SUPPORTED_FEATURES: print(f'{self._JSI_CLASS.__name__} does not support DOM, skipping') @@ -81,29 +98,32 @@ def test_execute_dom_script(self): if 'dom' not in self.jsi._SUPPORTED_FEATURES: print(f'{self._JSI_CLASS.__name__} does not support DOM, skipping') self.skipTest('DOM not supported') + self.assertEqual(self.jsi.execute( 'console.log(document.getElementById("test-div").innerHTML);', - html=''' + html='''Hello, world!
- + '''), 'Hello, world!') + def test_dom_location(self): + if not self._JSI_CLASS._SUPPORTED_FEATURES.issuperset({'dom', 'location'}): + print(f'{self._JSI_CLASS.__name__} does not support both DOM and location, skipping') + self.skipTest('DOM or location not supported') + + self.url_param = 'https://example.com/123/456' self.assertEqual(self.jsi.execute( 'console.log(document.getElementById("test-div").innerHTML);', - html=''' -
- - - '''), - 'Hello, world!') + html=''' +
Hello, world!
'''), + 'example.com') def test_execute_cookiejar(self): if 'cookies' not in self.jsi._SUPPORTED_FEATURES: @@ -134,7 +154,7 @@ def _assert_expected_execute(cookie_str, ref_cookie_str): ref_cookiejar.set_cookie(test_cookie.to_cookie()) # test identity without modification from js - self.url = 'http://example.com/123/456' + self.url_param = 'http://example.com/123/456' _assert_expected_execute(self.jsi.execute( 'console.log(document.cookie);', cookiejar=cookiejar), 'test1=test1; test3=test3') @@ -144,7 +164,7 @@ def _assert_expected_execute(cookie_str, ref_cookie_str): new_cookie_2 = NetscapeFields('test2', 'new2', '.example.com', '/', True, int(time.time()) + 900) ref_cookiejar.set_cookie(new_cookie_1.to_cookie()) ref_cookiejar.set_cookie(new_cookie_2.to_cookie()) - self.url = 'https://example.com/123/456' + self.url_param = 'https://example.com/123/456' _assert_expected_execute(self.jsi.execute( f'''document.cookie = "test1=new1; secure; expires={new_cookie_1.expire_str()}; domain=.example.com; path=/"; console.log(document.cookie);''', diff --git a/yt_dlp/jsinterp/_deno.py b/yt_dlp/jsinterp/_deno.py index f002881afd..8206f5b23f 100644 --- a/yt_dlp/jsinterp/_deno.py +++ b/yt_dlp/jsinterp/_deno.py @@ -142,7 +142,6 @@ def execute(self, jscode, video_id=None, note='Executing JS in Deno with jsdom', callback_varname = f'__callback_{random_string()}' script = f'''{self._init_script}; - {self._override_navigator_js}; import jsdom from "{self._JSDOM_URL}"; let {callback_varname} = (() => {{ const jar = jsdom.CookieJar.deserializeSync({json.dumps(self.serialize_cookie(cookiejar, self._url))}); @@ -151,9 +150,12 @@ def execute(self, jscode, video_id=None, note='Executing JS in Deno with jsdom', cookieJar: jar, pretendToBeVisual: true, }}); - Object.keys(dom.window).filter(key => !['atob', 'btoa'].includes(key)).forEach((key) => {{ - try {{window[key] = dom.window[key]}} catch (e) {{}} + Object.keys(dom.window).filter(key => !['atob', 'btoa', 'crypto', 'location'].includes(key)) + .filter(key => !(window.location? [] : ['sessionStorage', 'localStorage']).includes(key)) + .forEach((key) => {{ + try {{window[key] = dom.window[key]}} catch (e) {{ console.error(e) }} }}); + {self._override_navigator_js}; window.screen = {{ availWidth: 1920, @@ -168,8 +170,8 @@ def execute(self, jscode, video_id=None, note='Executing JS in Deno with jsdom', width: 1920, }} Object.defineProperty(document.body, 'clientWidth', {{value: 1903}}); - Object.defineProperty(document.body, 'clientHeight', {{value: 1035}}); - document.domain = location.hostname; + Object.defineProperty(document.body, 'clientHeight', {{value: 2000}}); + document.domain = location?.hostname; delete window.jsdom; const origLog = console.log; diff --git a/yt_dlp/jsinterp/_phantomjs.py b/yt_dlp/jsinterp/_phantomjs.py index e26d70c135..4c96ebeb04 100644 --- a/yt_dlp/jsinterp/_phantomjs.py +++ b/yt_dlp/jsinterp/_phantomjs.py @@ -164,26 +164,24 @@ def _execute_html(self, jscode: str, url: str, html: str, cookiejar, video_id=No return new_html, stdout def execute(self, jscode, video_id=None, note='Executing JS in PhantomJS', html='', cookiejar=None): - if self._url or html or cookiejar: - jscode = '''console.log(page.evaluate(function() { - var %(std_var)s = []; - console.log = function() { - var values = ''; - for (var i = 0; i < arguments.length; i++) { - values += arguments[i] + ' '; - } - %(std_var)s.push(values); + jscode = '''console.log(page.evaluate(function() { + var %(std_var)s = []; + console.log = function() { + var values = ''; + for (var i = 0; i < arguments.length; i++) { + values += arguments[i] + ' '; } - %(jscode)s; - return %(std_var)s.join('\\n'); - - })); - saveAndExit();''' % { - 'std_var': f'__stdout__values_{random_string()}', - 'jscode': jscode, + %(std_var)s.push(values); } - return self._execute_html(jscode, self._url, html, cookiejar, video_id=video_id, note=note)[1].strip() - return self._execute(jscode, video_id, note=note).strip() + %(jscode)s; + return %(std_var)s.join('\\n'); + + })); + saveAndExit();''' % { + 'std_var': f'__stdout__values_{random_string()}', + 'jscode': jscode, + } + return self._execute_html(jscode, self._url, html, cookiejar, video_id=video_id, note=note)[1].strip() class PhantomJSwrapper: diff --git a/yt_dlp/jsinterp/common.py b/yt_dlp/jsinterp/common.py index c2ef18ea9b..1a3bf00d19 100644 --- a/yt_dlp/jsinterp/common.py +++ b/yt_dlp/jsinterp/common.py @@ -117,7 +117,7 @@ def __init__( self.write_debug(f'Allowed JSI keys: {jsi_keys}') handler_classes = [_JSI_HANDLERS[key] for key in jsi_keys if _JSI_HANDLERS[key]._SUPPORTED_FEATURES.issuperset(self._features)] - self.write_debug(f'Selected JSI classes for given features: {get_jsi_keys(handler_classes)}, ' + self.write_debug(f'Select JSI for features={self._features}: {get_jsi_keys(handler_classes)}, ' f'included: {get_jsi_keys(only_include) or "all"}, excluded: {get_jsi_keys(exclude)}') self._handler_dict = {