[openload] Fix extraction (closes #16099)
[ytdl] / 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 (
12     compat_urlparse,
13     compat_kwargs,
14 )
15 from ..utils import (
16     check_executable,
17     determine_ext,
18     encodeArgument,
19     ExtractorError,
20     get_element_by_id,
21     get_exe_version,
22     is_outdated_version,
23     std_headers,
24 )
25
26
27 def cookie_to_dict(cookie):
28     cookie_dict = {
29         'name': cookie.name,
30         'value': cookie.value,
31     }
32     if cookie.port_specified:
33         cookie_dict['port'] = cookie.port
34     if cookie.domain_specified:
35         cookie_dict['domain'] = cookie.domain
36     if cookie.path_specified:
37         cookie_dict['path'] = cookie.path
38     if cookie.expires is not None:
39         cookie_dict['expires'] = cookie.expires
40     if cookie.secure is not None:
41         cookie_dict['secure'] = cookie.secure
42     if cookie.discard is not None:
43         cookie_dict['discard'] = cookie.discard
44     try:
45         if (cookie.has_nonstandard_attr('httpOnly') or
46                 cookie.has_nonstandard_attr('httponly') or
47                 cookie.has_nonstandard_attr('HttpOnly')):
48             cookie_dict['httponly'] = True
49     except TypeError:
50         pass
51     return cookie_dict
52
53
54 def cookie_jar_to_list(cookie_jar):
55     return [cookie_to_dict(cookie) for cookie in cookie_jar]
56
57
58 class PhantomJSwrapper(object):
59     """PhantomJS wrapper class
60
61     This class is experimental.
62     """
63
64     _TEMPLATE = r'''
65         phantom.onError = function(msg, trace) {{
66           var msgStack = ['PHANTOM ERROR: ' + msg];
67           if(trace && trace.length) {{
68             msgStack.push('TRACE:');
69             trace.forEach(function(t) {{
70               msgStack.push(' -> ' + (t.file || t.sourceURL) + ': ' + t.line
71                 + (t.function ? ' (in function ' + t.function +')' : ''));
72             }});
73           }}
74           console.error(msgStack.join('\n'));
75           phantom.exit(1);
76         }};
77         var page = require('webpage').create();
78         var fs = require('fs');
79         var read = {{ mode: 'r', charset: 'utf-8' }};
80         var write = {{ mode: 'w', charset: 'utf-8' }};
81         JSON.parse(fs.read("{cookies}", read)).forEach(function(x) {{
82           phantom.addCookie(x);
83         }});
84         page.settings.resourceTimeout = {timeout};
85         page.settings.userAgent = "{ua}";
86         page.onLoadStarted = function() {{
87           page.evaluate(function() {{
88             delete window._phantom;
89             delete window.callPhantom;
90           }});
91         }};
92         var saveAndExit = function() {{
93           fs.write("{html}", page.content, write);
94           fs.write("{cookies}", JSON.stringify(phantom.cookies), write);
95           phantom.exit();
96         }};
97         page.onLoadFinished = function(status) {{
98           if(page.url === "") {{
99             page.setContent(fs.read("{html}", read), "{url}");
100           }}
101           else {{
102             {jscode}
103           }}
104         }};
105         page.open("");
106     '''
107
108     _TMP_FILE_NAMES = ['script', 'html', 'cookies']
109
110     @staticmethod
111     def _version():
112         return get_exe_version('phantomjs', version_re=r'([0-9.]+)')
113
114     def __init__(self, extractor, required_version=None, timeout=10000):
115         self._TMP_FILES = {}
116
117         self.exe = check_executable('phantomjs', ['-v'])
118         if not self.exe:
119             raise ExtractorError('PhantomJS executable not found in PATH, '
120                                  'download it from http://phantomjs.org',
121                                  expected=True)
122
123         self.extractor = extractor
124
125         if required_version:
126             version = self._version()
127             if is_outdated_version(version, required_version):
128                 self.extractor._downloader.report_warning(
129                     'Your copy of PhantomJS is outdated, update it to version '
130                     '%s or newer if you encounter any errors.' % required_version)
131
132         self.options = {
133             'timeout': timeout,
134         }
135         for name in self._TMP_FILE_NAMES:
136             tmp = tempfile.NamedTemporaryFile(delete=False)
137             tmp.close()
138             self._TMP_FILES[name] = tmp
139
140     def __del__(self):
141         for name in self._TMP_FILE_NAMES:
142             try:
143                 os.remove(self._TMP_FILES[name].name)
144             except (IOError, OSError, KeyError):
145                 pass
146
147     def _save_cookies(self, url):
148         cookies = cookie_jar_to_list(self.extractor._downloader.cookiejar)
149         for cookie in cookies:
150             if 'path' not in cookie:
151                 cookie['path'] = '/'
152             if 'domain' not in cookie:
153                 cookie['domain'] = compat_urlparse.urlparse(url).netloc
154         with open(self._TMP_FILES['cookies'].name, 'wb') as f:
155             f.write(json.dumps(cookies).encode('utf-8'))
156
157     def _load_cookies(self):
158         with open(self._TMP_FILES['cookies'].name, 'rb') as f:
159             cookies = json.loads(f.read().decode('utf-8'))
160         for cookie in cookies:
161             if cookie['httponly'] is True:
162                 cookie['rest'] = {'httpOnly': None}
163             if 'expiry' in cookie:
164                 cookie['expire_time'] = cookie['expiry']
165             self.extractor._set_cookie(**compat_kwargs(cookie))
166
167     def get(self, url, html=None, video_id=None, note=None, note2='Executing JS on webpage', headers={}, jscode='saveAndExit();'):
168         """
169         Downloads webpage (if needed) and executes JS
170
171         Params:
172             url: website url
173             html: optional, html code of website
174             video_id: video id
175             note: optional, displayed when downloading webpage
176             note2: optional, displayed when executing JS
177             headers: custom http headers
178             jscode: code to be executed when page is loaded
179
180         Returns tuple with:
181             * downloaded website (after JS execution)
182             * anything you print with `console.log` (but not inside `page.execute`!)
183
184         In most cases you don't need to add any `jscode`.
185         It is executed in `page.onLoadFinished`.
186         `saveAndExit();` is mandatory, use it instead of `phantom.exit()`
187         It is possible to wait for some element on the webpage, for example:
188             var check = function() {
189               var elementFound = page.evaluate(function() {
190                 return document.querySelector('#b.done') !== null;
191               });
192               if(elementFound)
193                 saveAndExit();
194               else
195                 window.setTimeout(check, 500);
196             }
197
198             page.evaluate(function(){
199               document.querySelector('#a').click();
200             });
201             check();
202         """
203         if 'saveAndExit();' not in jscode:
204             raise ExtractorError('`saveAndExit();` not found in `jscode`')
205         if not html:
206             html = self.extractor._download_webpage(url, video_id, note=note, headers=headers)
207         with open(self._TMP_FILES['html'].name, 'wb') as f:
208             f.write(html.encode('utf-8'))
209
210         self._save_cookies(url)
211
212         replaces = self.options
213         replaces['url'] = url
214         user_agent = headers.get('User-Agent') or std_headers['User-Agent']
215         replaces['ua'] = user_agent.replace('"', '\\"')
216         replaces['jscode'] = jscode
217
218         for x in self._TMP_FILE_NAMES:
219             replaces[x] = self._TMP_FILES[x].name.replace('\\', '\\\\').replace('"', '\\"')
220
221         with open(self._TMP_FILES['script'].name, 'wb') as f:
222             f.write(self._TEMPLATE.format(**replaces).encode('utf-8'))
223
224         if video_id is None:
225             self.extractor.to_screen('%s' % (note2,))
226         else:
227             self.extractor.to_screen('%s: %s' % (video_id, note2))
228
229         p = subprocess.Popen([
230             self.exe, '--ssl-protocol=any',
231             self._TMP_FILES['script'].name
232         ], stdout=subprocess.PIPE, stderr=subprocess.PIPE)
233         out, err = p.communicate()
234         if p.returncode != 0:
235             raise ExtractorError(
236                 'Executing JS failed\n:' + encodeArgument(err))
237         with open(self._TMP_FILES['html'].name, 'rb') as f:
238             html = f.read().decode('utf-8')
239
240         self._load_cookies()
241
242         return (html, encodeArgument(out))
243
244
245 class OpenloadIE(InfoExtractor):
246     _VALID_URL = r'https?://(?:www\.)?(?:openload\.(?:co|io|link)|oload\.(?:tv|stream|site))/(?:f|embed)/(?P<id>[a-zA-Z0-9-_]+)'
247
248     _TESTS = [{
249         'url': 'https://openload.co/f/kUEfGclsU9o',
250         'md5': 'bf1c059b004ebc7a256f89408e65c36e',
251         'info_dict': {
252             'id': 'kUEfGclsU9o',
253             'ext': 'mp4',
254             'title': 'skyrim_no-audio_1080.mp4',
255             'thumbnail': r're:^https?://.*\.jpg$',
256         },
257     }, {
258         'url': 'https://openload.co/embed/rjC09fkPLYs',
259         'info_dict': {
260             'id': 'rjC09fkPLYs',
261             'ext': 'mp4',
262             'title': 'movie.mp4',
263             'thumbnail': r're:^https?://.*\.jpg$',
264             'subtitles': {
265                 'en': [{
266                     'ext': 'vtt',
267                 }],
268             },
269         },
270         'params': {
271             'skip_download': True,  # test subtitles only
272         },
273     }, {
274         'url': 'https://openload.co/embed/kUEfGclsU9o/skyrim_no-audio_1080.mp4',
275         'only_matching': True,
276     }, {
277         'url': 'https://openload.io/f/ZAn6oz-VZGE/',
278         'only_matching': True,
279     }, {
280         'url': 'https://openload.co/f/_-ztPaZtMhM/',
281         'only_matching': True,
282     }, {
283         # unavailable via https://openload.co/f/Sxz5sADo82g/, different layout
284         # for title and ext
285         'url': 'https://openload.co/embed/Sxz5sADo82g/',
286         'only_matching': True,
287     }, {
288         # unavailable via https://openload.co/embed/e-Ixz9ZR5L0/ but available
289         # via https://openload.co/f/e-Ixz9ZR5L0/
290         'url': 'https://openload.co/f/e-Ixz9ZR5L0/',
291         'only_matching': True,
292     }, {
293         'url': 'https://oload.tv/embed/KnG-kKZdcfY/',
294         'only_matching': True,
295     }, {
296         'url': 'http://www.openload.link/f/KnG-kKZdcfY',
297         'only_matching': True,
298     }, {
299         'url': 'https://oload.stream/f/KnG-kKZdcfY',
300         'only_matching': True,
301     }]
302
303     _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'
304
305     @staticmethod
306     def _extract_urls(webpage):
307         return re.findall(
308             r'<iframe[^>]+src=["\']((?:https?://)?(?:openload\.(?:co|io)|oload\.tv)/embed/[a-zA-Z0-9-_]+)',
309             webpage)
310
311     def _real_extract(self, url):
312         video_id = self._match_id(url)
313         url_pattern = 'https://openload.co/%%s/%s/' % video_id
314         headers = {
315             'User-Agent': self._USER_AGENT,
316         }
317
318         for path in ('embed', 'f'):
319             page_url = url_pattern % path
320             last = path == 'f'
321             webpage = self._download_webpage(
322                 page_url, video_id, 'Downloading %s webpage' % path,
323                 headers=headers, fatal=last)
324             if not webpage:
325                 continue
326             if 'File not found' in webpage or 'deleted by the owner' in webpage:
327                 if not last:
328                     continue
329                 raise ExtractorError('File not found', expected=True, video_id=video_id)
330             break
331
332         phantom = PhantomJSwrapper(self, required_version='2.0')
333         webpage, _ = phantom.get(page_url, html=webpage, video_id=video_id, headers=headers)
334
335         decoded_id = (get_element_by_id('streamurl', webpage) or
336                       get_element_by_id('streamuri', webpage) or
337                       get_element_by_id('streamurj', webpage) or
338                       self._search_regex(
339                           (r'>\s*([\da-zA-Z]+~\d{10,}~\d+\.\d+\.0\.0~[\da-zA-Z]+)\s*<',
340                            r'>\s*([\w~]+~\d+\.\d+\.\d+\.\d+~[\w~]+)'), webpage,
341                           'stream URL'))
342
343         video_url = 'https://openload.co/stream/%s?mime=true' % decoded_id
344
345         title = self._og_search_title(webpage, default=None) or self._search_regex(
346             r'<span[^>]+class=["\']title["\'][^>]*>([^<]+)', webpage,
347             'title', default=None) or self._html_search_meta(
348             'description', webpage, 'title', fatal=True)
349
350         entries = self._parse_html5_media_entries(page_url, webpage, video_id)
351         entry = entries[0] if entries else {}
352         subtitles = entry.get('subtitles')
353
354         info_dict = {
355             'id': video_id,
356             'title': title,
357             'thumbnail': entry.get('thumbnail') or self._og_search_thumbnail(webpage, default=None),
358             'url': video_url,
359             # Seems all videos have extensions in their titles
360             'ext': determine_ext(title, 'mp4'),
361             'subtitles': subtitles,
362             'http_headers': headers,
363         }
364         return info_dict