Merge pull request #14225 from Tithen-Firion/openload-phantomjs-method

Openload phantomjs method

youtube_dl/YoutubeDL.py

@@ -89,6 +89,7 @@ from .utils import (
     write_string,
     YoutubeDLCookieProcessor,
     YoutubeDLHandler,
+    PhantomJSwrapper,
 )
 from .cache import Cache
 from .extractor import get_info_extractor, gen_extractor_classes, _LAZY_LOADER
@@ -2217,6 +2218,7 @@ class YoutubeDL(object):
 
         exe_versions = FFmpegPostProcessor.get_versions(self)
         exe_versions['rtmpdump'] = rtmpdump_version()
+        exe_versions['phantomjs'] = PhantomJSwrapper._version()
         exe_str = ', '.join(
             '%s %s' % (exe, v)
             for exe, v in sorted(exe_versions.items())

youtube_dl/extractor/common.py

@@ -2449,10 +2449,12 @@ class InfoExtractor(object):
                 self._downloader.report_warning(msg)
         return res
 
-    def _set_cookie(self, domain, name, value, expire_time=None):
+    def _set_cookie(self, domain, name, value, expire_time=None, port=None,
+                    path='/', secure=False, discard=False, rest={}, **kwargs):
         cookie = compat_cookiejar.Cookie(
-            0, name, value, None, None, domain, None,
-            None, '/', True, False, expire_time, '', None, None, None)
+            0, name, value, port, not port is None, domain, True,
+            domain.startswith('.'), path, True, secure, expire_time,
+            discard, None, None, rest)
         self._downloader.cookiejar.set_cookie(cookie)
 
     def _get_cookies(self, url):

youtube_dl/extractor/openload.py

@@ -4,10 +4,11 @@ from __future__ import unicode_literals
 import re
 
 from .common import InfoExtractor
-from ..compat import compat_chr
 from ..utils import (
     determine_ext,
     ExtractorError,
+    get_element_by_id,
+    PhantomJSwrapper,
 )
 
 
@@ -58,6 +59,8 @@ class OpenloadIE(InfoExtractor):
         'only_matching': True,
     }]
 
+    _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'
+
     @staticmethod
     def _extract_urls(webpage):
         return re.findall(
@@ -66,47 +69,22 @@ class OpenloadIE(InfoExtractor):
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
-        webpage = self._download_webpage('https://openload.co/embed/%s/' % video_id, video_id)
+        url = 'https://openload.co/embed/%s/' % video_id
+        headers = {
+            'User-Agent': self._USER_AGENT,
+        }
+
+        webpage = self._download_webpage(url, video_id, headers=headers)
 
         if 'File not found' in webpage or 'deleted by the owner' in webpage:
-            raise ExtractorError('File not found', expected=True)
+            raise ExtractorError('File not found', expected=True, video_id=video_id)
 
-        ol_id = self._search_regex(
-            '<span[^>]+id="[^"]+"[^>]*>([0-9A-Za-z]+)</span>',
-            webpage, 'openload ID')
+        phantom = PhantomJSwrapper(self, required_version='2.0')
+        webpage, _ = phantom.get(url, html=webpage, video_id=video_id, headers=headers)
 
-        decoded = ''
-        a = ol_id[0:24]
-        b = []
-        for i in range(0, len(a), 8):
-            b.append(int(a[i:i + 8] or '0', 16))
-        ol_id = ol_id[24:]
-        j = 0
-        k = 0
-        while j < len(ol_id):
-            c = 128
-            d = 0
-            e = 0
-            f = 0
-            _more = True
-            while _more:
-                if j + 1 >= len(ol_id):
-                    c = 143
-                f = int(ol_id[j:j + 2] or '0', 16)
-                j += 2
-                d += (f & 127) << e
-                e += 7
-                _more = f >= c
-            g = d ^ b[k % 3]
-            for i in range(4):
-                char_dec = (g >> 8 * i) & (c + 127)
-                char = compat_chr(char_dec)
-                if char != '#':
-                    decoded += char
-            k += 1
+        decoded_id = get_element_by_id('streamurl', webpage)
 
-        video_url = 'https://openload.co/stream/%s?mime=true'
-        video_url = video_url % decoded
+        video_url = 'https://openload.co/stream/%s?mime=true' % decoded_id
 
         title = self._og_search_title(webpage, default=None) or self._search_regex(
             r'<span[^>]+class=["\']title["\'][^>]*>([^<]+)', webpage,
@@ -114,15 +92,17 @@ class OpenloadIE(InfoExtractor):
             'description', webpage, 'title', fatal=True)
 
         entries = self._parse_html5_media_entries(url, webpage, video_id)
-        subtitles = entries[0]['subtitles'] if entries else None
+        entry = entries[0] if entries else {}
+        subtitles = entry.get('subtitles')
 
         info_dict = {
             'id': video_id,
             'title': title,
-            'thumbnail': self._og_search_thumbnail(webpage, default=None),
+            'thumbnail': entry.get('thumbnail') or self._og_search_thumbnail(webpage, default=None),
             'url': video_url,
             # Seems all videos have extensions in their titles
             'ext': determine_ext(title, 'mp4'),
             'subtitles': subtitles,
+            'http_headers': headers,
         }
         return info_dict

youtube_dl/utils.py

@@ -3822,6 +3822,219 @@ 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"""
+
+    _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' }};
+        JSON.parse(fs.read("{cookies}", read)).forEach(function(x) {{
+          phantom.addCookie(x);
+        }});
+        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);
+          fs.write("{cookies}", JSON.stringify(phantom.cookies), 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', '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:
+            raise ExtractorError('PhantomJS executable not found in PATH, '
+                                 'download it from http://phantomjs.org',
+                                 expected=True)
+
+        self.extractor = extractor
+
+        if required_version:
+            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 '
+                    '%s or newer if you encounter any errors.' % required_version)
+
+        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 _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
+        
+        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'))
+
+        self._save_cookies(url)
+
+        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')
+
+        self._load_cookies()
+
+        return (html, encodeArgument(out))
+
+
 def random_birthday(year_field, month_field, day_field):
     return {
         year_field: str(random.randint(1950, 1995)),