[openload] separate PhantomJS code from extractor

youtube_dl/extractor/openload.py

@@ -1,17 +1,14 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
-import os
 import re
-import subprocess
-import tempfile
 
 from .common import InfoExtractor
 from ..utils import (
-    check_executable,
     determine_ext,
-    encodeArgument,
     ExtractorError,
+    get_element_by_id,
+    PhantomJSwrapper,
 )
 
 
@@ -62,38 +59,7 @@ class OpenloadIE(InfoExtractor):
         'only_matching': True,
     }]
 
-    _PHANTOMJS_SCRIPT = 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();
-        page.settings.resourceTimeout = 10000;
-        page.onInitialized = function() {
-          page.evaluate(function() {
-            delete window._phantom;
-            delete window.callPhantom;
-          });
-        };
-        page.open('https://openload.co/embed/%s/', function(status) {
-          var info = page.evaluate(function() {
-            return {
-              decoded_id: document.getElementById('streamurl').innerHTML,
-              title: document.querySelector('meta[name="og:title"],'
-                + 'meta[name=description]').content
-            };
-          });
-          console.log(info.decoded_id + ' ' + info.title);
-          phantom.exit();
-        });'''
+    _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):
@@ -102,40 +68,27 @@ def _extract_urls(webpage):
             webpage)
 
     def _real_extract(self, url):
-        exe = check_executable('phantomjs', ['-v'])
-        if not exe:
-            raise ExtractorError('PhantomJS executable not found in PATH, '
-                                 'download it from http://phantomjs.org',
-                                 expected=True)
-
         video_id = self._match_id(url)
         url = 'https://openload.co/embed/%s/' % video_id
-        webpage = self._download_webpage(url, video_id)
+        headers = {
+            'User-Agent': self._USER_AGENT,
+        }
+
+        phantom = PhantomJSwrapper(self)
+        webpage, _ = phantom.get(url, video_id=video_id, headers=headers)
 
         if 'File not found' in webpage or 'deleted by the owner' in webpage:
             raise ExtractorError('File not found', expected=True, video_id=video_id)
 
-        script_file = tempfile.NamedTemporaryFile(mode='w', delete=False)
-
-        # write JS script to file and close it
-        with script_file:
-            script_file.write(self._PHANTOMJS_SCRIPT % video_id)
-
-        self.to_screen('%s: Decoding video ID with PhantomJS' % video_id)
-
-        p = subprocess.Popen([exe, '--ssl-protocol=any', script_file.name],
-            stdout=subprocess.PIPE, stderr=subprocess.PIPE)
-        output, err = p.communicate()
-        if p.returncode != 0:
-            raise ExtractorError('Decoding failed\n:'
-                                 + encodeArgument(err))
-        else:
-            decoded_id, title = encodeArgument(output).strip().split(' ', 1)
-
-        os.remove(script_file.name)
+        decoded_id = get_element_by_id('streamurl', webpage)
 
         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,
+            'title', default=None) or self._html_search_meta(
+            'description', webpage, 'title', fatal=True)
+
         entries = self._parse_html5_media_entries(url, webpage, video_id)
         entry = entries[0] if entries else {}
         subtitles = entry.get('subtitles')
@@ -148,5 +101,6 @@ def _real_extract(self, 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

@@ -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))
+