]> Raphaël G. Git Repositories - youtubedl/blobdiff - youtube_dl/extractor/openload.py
Update README.md
[youtubedl] / youtube_dl / extractor / openload.py
index 6415b8fdcb7451e6b499d1d4d3515af2775c66c2..0c20d0177e0f2abbc5c0503babf51ba66715f374 100644 (file)
 # coding: utf-8
 from __future__ import unicode_literals
 
 # coding: utf-8
 from __future__ import unicode_literals
 
-import re
-
-from .common import InfoExtractor
-from ..compat import compat_chr
+import json
+import os
+import subprocess
+import tempfile
+
+from ..compat import (
+    compat_urlparse,
+    compat_kwargs,
+)
 from ..utils import (
 from ..utils import (
-    determine_ext,
-    encode_base_n,
+    check_executable,
+    encodeArgument,
     ExtractorError,
     ExtractorError,
-    mimetype2ext,
+    get_exe_version,
+    is_outdated_version,
+    std_headers,
 )
 
 
 )
 
 
-class OpenloadIE(InfoExtractor):
-    _VALID_URL = r'https://openload.(?:co|io)/(?:f|embed)/(?P<id>[a-zA-Z0-9-_]+)'
-
-    _TESTS = [{
-        'url': 'https://openload.co/f/kUEfGclsU9o',
-        'md5': 'bf1c059b004ebc7a256f89408e65c36e',
-        'info_dict': {
-            'id': 'kUEfGclsU9o',
-            'ext': 'mp4',
-            'title': 'skyrim_no-audio_1080.mp4',
-            'thumbnail': 're:^https?://.*\.jpg$',
-        },
-    }, {
-        'url': 'https://openload.co/embed/kUEfGclsU9o/skyrim_no-audio_1080.mp4',
-        'only_matching': True,
-    }, {
-        'url': 'https://openload.io/f/ZAn6oz-VZGE/',
-        'only_matching': True,
-    }, {
-        'url': 'https://openload.co/f/_-ztPaZtMhM/',
-        'only_matching': True,
-    }, {
-        # unavailable via https://openload.co/f/Sxz5sADo82g/, different layout
-        # for title and ext
-        'url': 'https://openload.co/embed/Sxz5sADo82g/',
-        'only_matching': True,
-    }]
+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 cookie.expires is not None:
+        cookie_dict['expires'] = cookie.expires
+    if cookie.secure is not None:
+        cookie_dict['secure'] = cookie.secure
+    if cookie.discard is not 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
+
+    This class is experimental.
+    """
+
+    _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
 
     @staticmethod
-    def openload_level2_debase(m):
-        radix, num = int(m.group(1)) + 27, int(m.group(2))
-        return '"' + encode_base_n(num, radix) + '"'
-
-    @classmethod
-    def openload_level2(cls, txt):
-        # The function name is ǃ \u01c3
-        # Using escaped unicode literals does not work in Python 3.2
-        return re.sub(r'ǃ\((\d+),(\d+)\)', cls.openload_level2_debase, txt, re.UNICODE).replace('"+"', '')
-
-    # Openload uses a variant of aadecode
-    # openload_decode and related functions are originally written by
-    # vitas@matfyz.cz and released with public domain
-    # See https://github.com/rg3/youtube-dl/issues/8489
-    @classmethod
-    def openload_decode(cls, txt):
-        symbol_table = [
-            ('_', '(゚Д゚) [゚Θ゚]'),
-            ('a', '(゚Д゚) [゚ω゚ノ]'),
-            ('b', '(゚Д゚) [゚Θ゚ノ]'),
-            ('c', '(゚Д゚) [\'c\']'),
-            ('d', '(゚Д゚) [゚ー゚ノ]'),
-            ('e', '(゚Д゚) [゚Д゚ノ]'),
-            ('f', '(゚Д゚) [1]'),
-
-            ('o', '(゚Д゚) [\'o\']'),
-            ('u', '(o゚ー゚o)'),
-            ('c', '(゚Д゚) [\'c\']'),
-
-            ('7', '((゚ー゚) + (o^_^o))'),
-            ('6', '((o^_^o) +(o^_^o) +(c^_^o))'),
-            ('5', '((゚ー゚) + (゚Θ゚))'),
-            ('4', '(-~3)'),
-            ('3', '(-~-~1)'),
-            ('2', '(-~1)'),
-            ('1', '(-~0)'),
-            ('0', '((c^_^o)-(c^_^o))'),
-        ]
-        delim = '(゚Д゚)[゚ε゚]+'
-        ret = ''
-        for aachar in txt.split(delim):
-            for val, pat in symbol_table:
-                aachar = aachar.replace(pat, val)
-            aachar = aachar.replace('+ ', '')
-            m = re.match(r'^\d+', aachar)
-            if m:
-                ret += compat_chr(int(m.group(0), 8))
-            else:
-                m = re.match(r'^u([\da-f]+)', aachar)
-                if m:
-                    ret += compat_chr(int(m.group(1), 16))
-        return cls.openload_level2(ret)
-
-    def _real_extract(self, url):
-        video_id = self._match_id(url)
-        webpage = self._download_webpage(url, video_id)
-
-        if 'File not found' in webpage:
-            raise ExtractorError('File not found', expected=True)
-
-        code = self._search_regex(
-            r'</video>\s*</div>\s*<script[^>]+>[^>]+</script>\s*<script[^>]+>([^<]+)</script>',
-            webpage, 'JS code')
-
-        decoded = self.openload_decode(code)
-
-        video_url = self._search_regex(
-            r'return\s+"(https?://[^"]+)"', decoded, 'video URL')
-
-        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)
-
-        ext = mimetype2ext(self._search_regex(
-            r'window\.vt\s*=\s*(["\'])(?P<mimetype>.+?)\1', decoded,
-            'mimetype', default=None, group='mimetype')) or determine_ext(
-            video_url, 'mp4')
-
-        return {
-            'id': video_id,
-            'title': title,
-            'ext': ext,
-            'thumbnail': self._og_search_thumbnail(webpage, default=None),
-            'url': video_url,
+    def _version():
+        return get_exe_version('phantomjs', version_re=r'([0-9.]+)')
+
+    def __init__(self, extractor, required_version=None, timeout=10000):
+        self._TMP_FILES = {}
+
+        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,
         }
         }
+        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 (IOError, OSError, KeyError):
+                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(**compat_kwargs(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))