From da57ebaf84225240b356530cdf02d12596f0dce8 Mon Sep 17 00:00:00 2001 From: Tithen-Firion Date: Tue, 25 Apr 2017 01:06:14 +0200 Subject: [openload] separate PhantomJS code from extractor --- youtube_dl/utils.py | 141 ++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 141 insertions(+) (limited to 'youtube_dl/utils.py') diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py index 2340bc306..94e1b07a6 100644 --- a/youtube_dl/utils.py +++ b/youtube_dl/utils.py @@ -3652,3 +3652,144 @@ def write_xattr(path, key, value): "Couldn't find a tool to set the xattrs. " "Install either the python 'xattr' module, " "or the 'xattr' binary.") + + +class PhantomJSwrapper(object): + """PhantomJS wrapper class""" + + _TEMPLATE = r''' + phantom.onError = function(msg, trace) {{ + var msgStack = ['PHANTOM ERROR: ' + msg]; + if(trace && trace.length) {{ + msgStack.push('TRACE:'); + trace.forEach(function(t) {{ + msgStack.push(' -> ' + (t.file || t.sourceURL) + ': ' + t.line + + (t.function ? ' (in function ' + t.function +')' : '')); + }}); + }} + console.error(msgStack.join('\n')); + phantom.exit(1); + }}; + var page = require('webpage').create(); + var fs = require('fs'); + var read = {{ mode: 'r', charset: 'utf-8' }}; + var write = {{ mode: 'w', charset: 'utf-8' }}; + page.settings.resourceTimeout = {timeout}; + page.settings.userAgent = "{ua}"; + page.onLoadStarted = function() {{ + page.evaluate(function() {{ + delete window._phantom; + delete window.callPhantom; + }}); + }}; + var saveAndExit = function() {{ + fs.write("{html}", page.content, write); + phantom.exit(); + }}; + page.onLoadFinished = function(status) {{ + if(page.url === "") {{ + page.setContent(fs.read("{html}", read), "{url}"); + }} + else {{ + {jscode} + }} + }}; + page.open(""); + ''' + + _TMP_FILE_NAMES = ['script', 'html'] + + def __init__(self, extractor, timeout=10000): + self.exe = check_executable('phantomjs', ['-v']) + if not self.exe: + raise ExtractorError('PhantomJS executable not found in PATH, ' + 'download it from http://phantomjs.org', + expected=True) + self.extractor = extractor + self.options = { + 'timeout': timeout, + } + self._TMP_FILES = {} + for name in self._TMP_FILE_NAMES: + tmp = tempfile.NamedTemporaryFile(delete=False) + tmp.close() + self._TMP_FILES[name] = tmp + + def __del__(self): + for name in self._TMP_FILE_NAMES: + try: + os.remove(self._TMP_FILES[name].name) + except: + pass + + def get(self, url, html=None, video_id=None, note=None, note2='Executing JS on webpage', headers={}, jscode='saveAndExit();'): + """ + Downloads webpage (if needed) and executes JS + + Params: + url: website url + html: optional, html code of website + video_id: video id + note: optional, displayed when downloading webpage + note2: optional, displayed when executing JS + headers: custom http headers + jscode: code to be executed when page is loaded + + Returns tuple with: + * downloaded website (after JS execution) + * anything you print with `console.log` (but not inside `page.execute`!) + + In most cases you don't need to add any `jscode`. + It is executed in `page.onLoadFinished`. + `saveAndExit();` is mandatory, use it instead of `phantom.exit()` + It is possible to wait for some element on the webpage, for example: + var check = function() { + var elementFound = page.evaluate(function() { + return document.querySelector('#b.done') !== null; + }); + if(elementFound) + saveAndExit(); + else + window.setTimeout(check, 500); + } + + page.evaluate(function(){ + document.querySelector('#a').click(); + }); + check(); + """ + if 'saveAndExit();' not in jscode: + raise ExtractorError('`saveAndExit();` not found in `jscode`') + if not html: + html = self.extractor._download_webpage(url, video_id, note=note, headers=headers) + with open(self._TMP_FILES['html'].name, 'wb') as f: + f.write(html.encode('utf-8')) + + replaces = self.options + replaces['url'] = url + user_agent = headers.get('User-Agent') or std_headers['User-Agent'] + replaces['ua'] = user_agent.replace('"', '\\"') + replaces['jscode'] = jscode + + for x in self._TMP_FILE_NAMES: + replaces[x] = self._TMP_FILES[x].name.replace('\\', '\\\\').replace('"', '\\"') + + with open(self._TMP_FILES['script'].name, 'wb') as f: + f.write(self._TEMPLATE.format(**replaces).encode('utf-8')) + + if video_id is None: + self.extractor.to_screen('%s' % (note2,)) + else: + self.extractor.to_screen('%s: %s' % (video_id, note2)) + + p = subprocess.Popen([self.exe, '--ssl-protocol=any', + self._TMP_FILES['script'].name], stdout=subprocess.PIPE, + stderr=subprocess.PIPE) + out, err = p.communicate() + if p.returncode != 0: + raise ExtractorError('Executing JS failed\n:' + + encodeArgument(err)) + with open(self._TMP_FILES['html'].name, 'rb') as f: + html = f.read().decode('utf-8') + return (html, encodeArgument(out)) + -- cgit v1.2.3 From 40e41780f1d770a355f01e3c1e6fb09ff392f97e Mon Sep 17 00:00:00 2001 From: Tithen-Firion Date: Tue, 25 Apr 2017 15:12:54 +0200 Subject: [phantomjs] add cookie support --- youtube_dl/utils.py | 62 ++++++++++++++++++++++++++++++++++++++++++++++++++++- 1 file changed, 61 insertions(+), 1 deletion(-) (limited to 'youtube_dl/utils.py') diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py index 94e1b07a6..9c94b7ec9 100644 --- a/youtube_dl/utils.py +++ b/youtube_dl/utils.py @@ -3654,6 +3654,37 @@ def write_xattr(path, key, value): "or the 'xattr' binary.") +def cookie_to_dict(cookie): + cookie_dict = { + 'name': cookie.name, + 'value': cookie.value, + }; + if cookie.port_specified: + cookie_dict['port'] = cookie.port + if cookie.domain_specified: + cookie_dict['domain'] = cookie.domain + if cookie.path_specified: + cookie_dict['path'] = cookie.path + if not cookie.expires is None: + cookie_dict['expires'] = cookie.expires + if not cookie.secure is None: + cookie_dict['secure'] = cookie.secure + if not cookie.discard is None: + cookie_dict['discard'] = cookie.discard + try: + if (cookie.has_nonstandard_attr('httpOnly') or + cookie.has_nonstandard_attr('httponly') or + cookie.has_nonstandard_attr('HttpOnly')): + cookie_dict['httponly'] = True + except TypeError: + pass + return cookie_dict + + +def cookie_jar_to_list(cookie_jar): + return [cookie_to_dict(cookie) for cookie in cookie_jar] + + class PhantomJSwrapper(object): """PhantomJS wrapper class""" @@ -3674,6 +3705,9 @@ class PhantomJSwrapper(object): var fs = require('fs'); var read = {{ mode: 'r', charset: 'utf-8' }}; var write = {{ mode: 'w', charset: 'utf-8' }}; + JSON.parse(fs.read("{cookies}", read)).forEach(function(x) {{ + phantom.addCookie(x); + }}); page.settings.resourceTimeout = {timeout}; page.settings.userAgent = "{ua}"; page.onLoadStarted = function() {{ @@ -3684,6 +3718,7 @@ class PhantomJSwrapper(object): }}; var saveAndExit = function() {{ fs.write("{html}", page.content, write); + fs.write("{cookies}", JSON.stringify(phantom.cookies), write); phantom.exit(); }}; page.onLoadFinished = function(status) {{ @@ -3697,7 +3732,7 @@ class PhantomJSwrapper(object): page.open(""); ''' - _TMP_FILE_NAMES = ['script', 'html'] + _TMP_FILE_NAMES = ['script', 'html', 'cookies'] def __init__(self, extractor, timeout=10000): self.exe = check_executable('phantomjs', ['-v']) @@ -3722,6 +3757,26 @@ class PhantomJSwrapper(object): except: pass + def _save_cookies(self, url): + cookies = cookie_jar_to_list(self.extractor._downloader.cookiejar) + for cookie in cookies: + if 'path' not in cookie: + cookie['path'] = '/' + if 'domain' not in cookie: + cookie['domain'] = compat_urlparse.urlparse(url).netloc + with open(self._TMP_FILES['cookies'].name, 'wb') as f: + f.write(json.dumps(cookies).encode('utf-8')) + + def _load_cookies(self): + with open(self._TMP_FILES['cookies'].name, 'rb') as f: + cookies = json.loads(f.read().decode('utf-8')) + for cookie in cookies: + if cookie['httponly'] is True: + cookie['rest'] = { 'httpOnly': None } + if 'expiry' in cookie: + cookie['expire_time'] = cookie['expiry'] + self.extractor._set_cookie(**cookie) + def get(self, url, html=None, video_id=None, note=None, note2='Executing JS on webpage', headers={}, jscode='saveAndExit();'): """ Downloads webpage (if needed) and executes JS @@ -3765,6 +3820,8 @@ class PhantomJSwrapper(object): with open(self._TMP_FILES['html'].name, 'wb') as f: f.write(html.encode('utf-8')) + self._save_cookies(url) + replaces = self.options replaces['url'] = url user_agent = headers.get('User-Agent') or std_headers['User-Agent'] @@ -3791,5 +3848,8 @@ class PhantomJSwrapper(object): + encodeArgument(err)) with open(self._TMP_FILES['html'].name, 'rb') as f: html = f.read().decode('utf-8') + + self._load_cookies() + return (html, encodeArgument(out)) -- cgit v1.2.3 From 98f9d873814da2a8584cc30c0e197c15ed249db3 Mon Sep 17 00:00:00 2001 From: Tithen-Firion Date: Sat, 29 Apr 2017 12:41:42 +0200 Subject: [phantomjs] Add required version checking --- youtube_dl/utils.py | 11 ++++++++++- 1 file changed, 10 insertions(+), 1 deletion(-) (limited to 'youtube_dl/utils.py') diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py index 9c94b7ec9..84aaac664 100644 --- a/youtube_dl/utils.py +++ b/youtube_dl/utils.py @@ -3734,13 +3734,22 @@ class PhantomJSwrapper(object): _TMP_FILE_NAMES = ['script', 'html', 'cookies'] - def __init__(self, extractor, timeout=10000): + def __init__(self, extractor, required_version=None, timeout=10000): self.exe = check_executable('phantomjs', ['-v']) if not self.exe: raise ExtractorError('PhantomJS executable not found in PATH, ' 'download it from http://phantomjs.org', expected=True) + self.extractor = extractor + + if required_version: + version = get_exe_version(self.exe, version_re=r'([0-9.]+)') + if is_outdated_version(version, required_version): + self.extractor._downloader.report_warning( + 'Your copy of PhantomJS is outdated, update it to version ' + '%s or newer if you encounter any errors.' % required_version) + self.options = { 'timeout': timeout, } -- cgit v1.2.3 From feee8d32e45c9521426cf4a089c70f37542f0065 Mon Sep 17 00:00:00 2001 From: Tithen-Firion Date: Thu, 3 Aug 2017 14:17:25 +0200 Subject: [phantomjs] add exe version to debug info --- youtube_dl/utils.py | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) (limited to 'youtube_dl/utils.py') diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py index c67f95ac9..4d0685d83 100644 --- a/youtube_dl/utils.py +++ b/youtube_dl/utils.py @@ -3895,6 +3895,10 @@ class PhantomJSwrapper(object): _TMP_FILE_NAMES = ['script', 'html', 'cookies'] + @staticmethod + def _version(): + return get_exe_version('phantomjs', version_re=r'([0-9.]+)') + def __init__(self, extractor, required_version=None, timeout=10000): self.exe = check_executable('phantomjs', ['-v']) if not self.exe: @@ -3905,7 +3909,7 @@ class PhantomJSwrapper(object): self.extractor = extractor if required_version: - version = get_exe_version(self.exe, version_re=r'([0-9.]+)') + version = self._version() if is_outdated_version(version, required_version): self.extractor._downloader.report_warning( 'Your copy of PhantomJS is outdated, update it to version ' -- cgit v1.2.3