_ Git - youtube-dl/blob - youtube_dl/extractor/openload.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import json
   5 import os
   6 import re
   7 import subprocess
   8 import tempfile
   9
  10 from .common import InfoExtractor
  11 from ..compat import compat_urlparse
  12 from ..utils import (
  13     check_executable,
  14     determine_ext,
  15     encodeArgument,
  16     ExtractorError,
  17     get_element_by_id,
  18     get_exe_version,
  19     is_outdated_version,
  20     std_headers,
  21 )
  22
  23
  24 def cookie_to_dict(cookie):
  25     cookie_dict = {
  26         'name': cookie.name,
  27         'value': cookie.value,
  28     }
  29     if cookie.port_specified:
  30         cookie_dict['port'] = cookie.port
  31     if cookie.domain_specified:
  32         cookie_dict['domain'] = cookie.domain
  33     if cookie.path_specified:
  34         cookie_dict['path'] = cookie.path
  35     if cookie.expires is not None:
  36         cookie_dict['expires'] = cookie.expires
  37     if cookie.secure is not None:
  38         cookie_dict['secure'] = cookie.secure
  39     if cookie.discard is not None:
  40         cookie_dict['discard'] = cookie.discard
  41     try:
  42         if (cookie.has_nonstandard_attr('httpOnly') or
  43                 cookie.has_nonstandard_attr('httponly') or
  44                 cookie.has_nonstandard_attr('HttpOnly')):
  45             cookie_dict['httponly'] = True
  46     except TypeError:
  47         pass
  48     return cookie_dict
  49
  50
  51 def cookie_jar_to_list(cookie_jar):
  52     return [cookie_to_dict(cookie) for cookie in cookie_jar]
  53
  54
  55 class PhantomJSwrapper(object):
  56     """PhantomJS wrapper class
  57
  58     This class is experimental.
  59     """
  60
  61     _TEMPLATE = r'''
  62         phantom.onError = function(msg, trace) {{
  63           var msgStack = ['PHANTOM ERROR: ' + msg];
  64           if(trace && trace.length) {{
  65             msgStack.push('TRACE:');
  66             trace.forEach(function(t) {{
  67               msgStack.push(' -> ' + (t.file || t.sourceURL) + ': ' + t.line
  68                 + (t.function ? ' (in function ' + t.function +')' : ''));
  69             }});
  70           }}
  71           console.error(msgStack.join('\n'));
  72           phantom.exit(1);
  73         }};
  74         var page = require('webpage').create();
  75         var fs = require('fs');
  76         var read = {{ mode: 'r', charset: 'utf-8' }};
  77         var write = {{ mode: 'w', charset: 'utf-8' }};
  78         JSON.parse(fs.read("{cookies}", read)).forEach(function(x) {{
  79           phantom.addCookie(x);
  80         }});
  81         page.settings.resourceTimeout = {timeout};
  82         page.settings.userAgent = "{ua}";
  83         page.onLoadStarted = function() {{
  84           page.evaluate(function() {{
  85             delete window._phantom;
  86             delete window.callPhantom;
  87           }});
  88         }};
  89         var saveAndExit = function() {{
  90           fs.write("{html}", page.content, write);
  91           fs.write("{cookies}", JSON.stringify(phantom.cookies), write);
  92           phantom.exit();
  93         }};
  94         page.onLoadFinished = function(status) {{
  95           if(page.url === "") {{
  96             page.setContent(fs.read("{html}", read), "{url}");
  97           }}
  98           else {{
  99             {jscode}
 100           }}
 101         }};
 102         page.open("");
 103     '''
 104
 105     _TMP_FILE_NAMES = ['script', 'html', 'cookies']
 106
 107     @staticmethod
 108     def _version():
 109         return get_exe_version('phantomjs', version_re=r'([0-9.]+)')
 110
 111     def __init__(self, extractor, required_version=None, timeout=10000):
 112         self.exe = check_executable('phantomjs', ['-v'])
 113         if not self.exe:
 114             raise ExtractorError('PhantomJS executable not found in PATH, '
 115                                  'download it from http://phantomjs.org',
 116                                  expected=True)
 117
 118         self.extractor = extractor
 119
 120         if required_version:
 121             version = self._version()
 122             if is_outdated_version(version, required_version):
 123                 self.extractor._downloader.report_warning(
 124                     'Your copy of PhantomJS is outdated, update it to version '
 125                     '%s or newer if you encounter any errors.' % required_version)
 126
 127         self.options = {
 128             'timeout': timeout,
 129         }
 130         self._TMP_FILES = {}
 131         for name in self._TMP_FILE_NAMES:
 132             tmp = tempfile.NamedTemporaryFile(delete=False)
 133             tmp.close()
 134             self._TMP_FILES[name] = tmp
 135
 136     def __del__(self):
 137         for name in self._TMP_FILE_NAMES:
 138             try:
 139                 os.remove(self._TMP_FILES[name].name)
 140             except:
 141                 pass
 142
 143     def _save_cookies(self, url):
 144         cookies = cookie_jar_to_list(self.extractor._downloader.cookiejar)
 145         for cookie in cookies:
 146             if 'path' not in cookie:
 147                 cookie['path'] = '/'
 148             if 'domain' not in cookie:
 149                 cookie['domain'] = compat_urlparse.urlparse(url).netloc
 150         with open(self._TMP_FILES['cookies'].name, 'wb') as f:
 151             f.write(json.dumps(cookies).encode('utf-8'))
 152
 153     def _load_cookies(self):
 154         with open(self._TMP_FILES['cookies'].name, 'rb') as f:
 155             cookies = json.loads(f.read().decode('utf-8'))
 156         for cookie in cookies:
 157             if cookie['httponly'] is True:
 158                 cookie['rest'] = {'httpOnly': None}
 159             if 'expiry' in cookie:
 160                 cookie['expire_time'] = cookie['expiry']
 161             self.extractor._set_cookie(**cookie)
 162
 163     def get(self, url, html=None, video_id=None, note=None, note2='Executing JS on webpage', headers={}, jscode='saveAndExit();'):
 164         """
 165         Downloads webpage (if needed) and executes JS
 166
 167         Params:
 168             url: website url
 169             html: optional, html code of website
 170             video_id: video id
 171             note: optional, displayed when downloading webpage
 172             note2: optional, displayed when executing JS
 173             headers: custom http headers
 174             jscode: code to be executed when page is loaded
 175
 176         Returns tuple with:
 177             * downloaded website (after JS execution)
 178             * anything you print with `console.log` (but not inside `page.execute`!)
 179
 180         In most cases you don't need to add any `jscode`.
 181         It is executed in `page.onLoadFinished`.
 182         `saveAndExit();` is mandatory, use it instead of `phantom.exit()`
 183         It is possible to wait for some element on the webpage, for example:
 184             var check = function() {
 185               var elementFound = page.evaluate(function() {
 186                 return document.querySelector('#b.done') !== null;
 187               });
 188               if(elementFound)
 189                 saveAndExit();
 190               else
 191                 window.setTimeout(check, 500);
 192             }
 193
 194             page.evaluate(function(){
 195               document.querySelector('#a').click();
 196             });
 197             check();
 198         """
 199         if 'saveAndExit();' not in jscode:
 200             raise ExtractorError('`saveAndExit();` not found in `jscode`')
 201         if not html:
 202             html = self.extractor._download_webpage(url, video_id, note=note, headers=headers)
 203         with open(self._TMP_FILES['html'].name, 'wb') as f:
 204             f.write(html.encode('utf-8'))
 205
 206         self._save_cookies(url)
 207
 208         replaces = self.options
 209         replaces['url'] = url
 210         user_agent = headers.get('User-Agent') or std_headers['User-Agent']
 211         replaces['ua'] = user_agent.replace('"', '\\"')
 212         replaces['jscode'] = jscode
 213
 214         for x in self._TMP_FILE_NAMES:
 215             replaces[x] = self._TMP_FILES[x].name.replace('\\', '\\\\').replace('"', '\\"')
 216
 217         with open(self._TMP_FILES['script'].name, 'wb') as f:
 218             f.write(self._TEMPLATE.format(**replaces).encode('utf-8'))
 219
 220         if video_id is None:
 221             self.extractor.to_screen('%s' % (note2,))
 222         else:
 223             self.extractor.to_screen('%s: %s' % (video_id, note2))
 224
 225         p = subprocess.Popen([
 226             self.exe, '--ssl-protocol=any',
 227             self._TMP_FILES['script'].name
 228         ], stdout=subprocess.PIPE, stderr=subprocess.PIPE)
 229         out, err = p.communicate()
 230         if p.returncode != 0:
 231             raise ExtractorError(
 232                 'Executing JS failed\n:' + encodeArgument(err))
 233         with open(self._TMP_FILES['html'].name, 'rb') as f:
 234             html = f.read().decode('utf-8')
 235
 236         self._load_cookies()
 237
 238         return (html, encodeArgument(out))
 239
 240
 241 class OpenloadIE(InfoExtractor):
 242     _VALID_URL = r'https?://(?:openload\.(?:co|io)|oload\.tv)/(?:f|embed)/(?P<id>[a-zA-Z0-9-_]+)'
 243
 244     _TESTS = [{
 245         'url': 'https://openload.co/f/kUEfGclsU9o',
 246         'md5': 'bf1c059b004ebc7a256f89408e65c36e',
 247         'info_dict': {
 248             'id': 'kUEfGclsU9o',
 249             'ext': 'mp4',
 250             'title': 'skyrim_no-audio_1080.mp4',
 251             'thumbnail': r're:^https?://.*\.jpg$',
 252         },
 253     }, {
 254         'url': 'https://openload.co/embed/rjC09fkPLYs',
 255         'info_dict': {
 256             'id': 'rjC09fkPLYs',
 257             'ext': 'mp4',
 258             'title': 'movie.mp4',
 259             'thumbnail': r're:^https?://.*\.jpg$',
 260             'subtitles': {
 261                 'en': [{
 262                     'ext': 'vtt',
 263                 }],
 264             },
 265         },
 266         'params': {
 267             'skip_download': True,  # test subtitles only
 268         },
 269     }, {
 270         'url': 'https://openload.co/embed/kUEfGclsU9o/skyrim_no-audio_1080.mp4',
 271         'only_matching': True,
 272     }, {
 273         'url': 'https://openload.io/f/ZAn6oz-VZGE/',
 274         'only_matching': True,
 275     }, {
 276         'url': 'https://openload.co/f/_-ztPaZtMhM/',
 277         'only_matching': True,
 278     }, {
 279         # unavailable via https://openload.co/f/Sxz5sADo82g/, different layout
 280         # for title and ext
 281         'url': 'https://openload.co/embed/Sxz5sADo82g/',
 282         'only_matching': True,
 283     }, {
 284         'url': 'https://oload.tv/embed/KnG-kKZdcfY/',
 285         'only_matching': True,
 286     }]
 287
 288     _USER_AGENT = 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/56.0.2924.87 Safari/537.36'
 289
 290     @staticmethod
 291     def _extract_urls(webpage):
 292         return re.findall(
 293             r'<iframe[^>]+src=["\']((?:https?://)?(?:openload\.(?:co|io)|oload\.tv)/embed/[a-zA-Z0-9-_]+)',
 294             webpage)
 295
 296     def _real_extract(self, url):
 297         video_id = self._match_id(url)
 298         url = 'https://openload.co/embed/%s/' % video_id
 299         headers = {
 300             'User-Agent': self._USER_AGENT,
 301         }
 302
 303         webpage = self._download_webpage(url, video_id, headers=headers)
 304
 305         if 'File not found' in webpage or 'deleted by the owner' in webpage:
 306             raise ExtractorError('File not found', expected=True, video_id=video_id)
 307
 308         phantom = PhantomJSwrapper(self, required_version='2.0')
 309         webpage, _ = phantom.get(url, html=webpage, video_id=video_id, headers=headers)
 310
 311         decoded_id = get_element_by_id('streamurl', webpage)
 312
 313         video_url = 'https://openload.co/stream/%s?mime=true' % decoded_id
 314
 315         title = self._og_search_title(webpage, default=None) or self._search_regex(
 316             r'<span[^>]+class=["\']title["\'][^>]*>([^<]+)', webpage,
 317             'title', default=None) or self._html_search_meta(
 318             'description', webpage, 'title', fatal=True)
 319
 320         entries = self._parse_html5_media_entries(url, webpage, video_id)
 321         entry = entries[0] if entries else {}
 322         subtitles = entry.get('subtitles')
 323
 324         info_dict = {
 325             'id': video_id,
 326             'title': title,
 327             'thumbnail': entry.get('thumbnail') or self._og_search_thumbnail(webpage, default=None),
 328             'url': video_url,
 329             # Seems all videos have extensions in their titles
 330             'ext': determine_ext(title, 'mp4'),
 331             'subtitles': subtitles,
 332             'http_headers': headers,
 333         }
 334         return info_dict