-# coding: utf-8
-from __future__ import unicode_literals
-
+import collections
+import contextlib
import json
import os
import subprocess
import tempfile
+import urllib.parse
-from ..compat import (
- compat_urlparse,
-)
from ..utils import (
- check_executable,
- encodeArgument,
ExtractorError,
+ Popen,
+ check_executable,
+ format_field,
get_exe_version,
is_outdated_version,
- Popen,
+ shell_quote,
)
cookie_dict['secure'] = cookie.secure
if cookie.discard is not None:
cookie_dict['discard'] = cookie.discard
- try:
+ with contextlib.suppress(TypeError):
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
return [cookie_to_dict(cookie) for cookie in cookie_jar]
-class PhantomJSwrapper(object):
+class PhantomJSwrapper:
"""PhantomJS wrapper class
This class is experimental.
"""
- _TEMPLATE = r'''
+ INSTALL_HINT = 'Please download it from https://phantomjs.org/download.html'
+
+ _BASE_JS = R'''
phantom.onError = function(msg, trace) {{
var msgStack = ['PHANTOM ERROR: ' + msg];
if(trace && trace.length) {{
console.error(msgStack.join('\n'));
phantom.exit(1);
}};
+ '''
+
+ _TEMPLATE = R'''
var page = require('webpage').create();
var fs = require('fs');
var read = {{ mode: 'r', charset: 'utf-8' }};
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)
+ raise ExtractorError(f'PhantomJS not found, {self.INSTALL_HINT}', expected=True)
self.extractor = extractor
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'Your copy of PhantomJS is outdated, update it to version '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)
+ f'{required_version} or newer if you encounter any errors.')
- self.options = {
- 'timeout': timeout,
- }
for name in self._TMP_FILE_NAMES:
tmp = tempfile.NamedTemporaryFile(delete=False)
tmp.close()
self._TMP_FILES[name] = tmp
+ self.options = collections.ChainMap({
+ 'timeout': timeout,
+ }, {
+ x: self._TMP_FILES[x].name.replace('\\', '\\\\').replace('"', '\\"')
+ for x in self._TMP_FILE_NAMES
+ })
+
def __del__(self):
for name in self._TMP_FILE_NAMES:
- try:
+ with contextlib.suppress(OSError, KeyError):
os.remove(self._TMP_FILES[name].name)
- except (IOError, OSError, KeyError):
- pass
def _save_cookies(self, url):
- cookies = cookie_jar_to_list(self.extractor._downloader.cookiejar)
+ cookies = cookie_jar_to_list(self.extractor.cookiejar)
for cookie in cookies:
if 'path' not in cookie:
cookie['path'] = '/'
if 'domain' not in cookie:
- cookie['domain'] = compat_urlparse.urlparse(url).netloc
+ cookie['domain'] = urllib.parse.urlparse(url).netloc
with open(self._TMP_FILES['cookies'].name, 'wb') as f:
- f.write(json.dumps(cookies).encode('utf-8'))
+ f.write(json.dumps(cookies).encode())
def _load_cookies(self):
with open(self._TMP_FILES['cookies'].name, 'rb') as f:
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:
+ It is possible to wait for some element on the webpage, e.g.
var check = function() {
var elementFound = page.evaluate(function() {
return document.querySelector('#b.done') !== null;
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'))
+ f.write(html.encode())
self._save_cookies(url)
- replaces = self.options
- replaces['url'] = url
user_agent = headers.get('User-Agent') or self.extractor.get_param('http_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 = Popen(
- [self.exe, '--ssl-protocol=any', self._TMP_FILES['script'].name],
- stdout=subprocess.PIPE, stderr=subprocess.PIPE)
- out, err = p.communicate_or_kill()
- if p.returncode != 0:
- raise ExtractorError(
- 'Executing JS failed\n:' + encodeArgument(err))
+ jscode = self._TEMPLATE.format_map(self.options.new_child({
+ 'url': url,
+ 'ua': user_agent.replace('"', '\\"'),
+ 'jscode': jscode,
+ }))
+
+ stdout = self.execute(jscode, video_id, note=note2)
+
with open(self._TMP_FILES['html'].name, 'rb') as f:
html = f.read().decode('utf-8')
-
self._load_cookies()
- return (html, encodeArgument(out))
+ return html, stdout
+
+ def execute(self, jscode, video_id=None, *, note='Executing JS'):
+ """Execute JS and return stdout"""
+ if 'phantom.exit();' not in jscode:
+ jscode += ';\nphantom.exit();'
+ jscode = self._BASE_JS + jscode
+
+ with open(self._TMP_FILES['script'].name, 'w', encoding='utf-8') as f:
+ f.write(jscode)
+ self.extractor.to_screen(f'{format_field(video_id, None, "%s: ")}{note}')
+
+ cmd = [self.exe, '--ssl-protocol=any', self._TMP_FILES['script'].name]
+ self.extractor.write_debug(f'PhantomJS command line: {shell_quote(cmd)}')
+ try:
+ stdout, stderr, returncode = Popen.run(cmd, timeout=self.options['timeout'] / 1000,
+ text=True, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
+ except Exception as e:
+ raise ExtractorError(f'{note} failed: Unable to run PhantomJS binary', cause=e)
+ if returncode:
+ raise ExtractorError(f'{note} failed with returncode {returncode}:\n{stderr.strip()}')
+
+ return stdout