Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/facebook.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import re
   5 import socket
   6
   7 from .common import InfoExtractor
   8 from ..compat import (
   9     compat_etree_fromstring,
  10     compat_http_client,
  11     compat_urllib_error,
  12     compat_urllib_parse_unquote,
  13     compat_urllib_parse_unquote_plus,
  14 )
  15 from ..utils import (
  16     clean_html,
  17     error_to_compat_str,
  18     ExtractorError,
  19     get_element_by_id,
  20     int_or_none,
  21     js_to_json,
  22     limit_length,
  23     sanitized_Request,
  24     try_get,
  25     urlencode_postdata,
  26 )
  27
  28
  29 class FacebookIE(InfoExtractor):
  30     _VALID_URL = r'''(?x)
  31                 (?:
  32                     https?://
  33                         (?:[\w-]+\.)?(?:facebook\.com|facebookcorewwwi\.onion)/
  34                         (?:[^#]*?\#!/)?
  35                         (?:
  36                             (?:
  37                                 video/video\.php|
  38                                 photo\.php|
  39                                 video\.php|
  40                                 video/embed|
  41                                 story\.php
  42                             )\?(?:.*?)(?:v|video_id|story_fbid)=|
  43                             [^/]+/videos/(?:[^/]+/)?|
  44                             [^/]+/posts/|
  45                             groups/[^/]+/permalink/
  46                         )|
  47                     facebook:
  48                 )
  49                 (?P<id>[0-9]+)
  50                 '''
  51     _LOGIN_URL = 'https://www.facebook.com/login.php?next=http%3A%2F%2Ffacebook.com%2Fhome.php&login_attempt=1'
  52     _CHECKPOINT_URL = 'https://www.facebook.com/checkpoint/?next=http%3A%2F%2Ffacebook.com%2Fhome.php&_fb_noscript=1'
  53     _NETRC_MACHINE = 'facebook'
  54     IE_NAME = 'facebook'
  55
  56     _CHROME_USER_AGENT = 'Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/48.0.2564.97 Safari/537.36'
  57
  58     _VIDEO_PAGE_TEMPLATE = 'https://www.facebook.com/video/video.php?v=%s'
  59
  60     _TESTS = [{
  61         'url': 'https://www.facebook.com/video.php?v=637842556329505&fref=nf',
  62         'md5': '6a40d33c0eccbb1af76cf0485a052659',
  63         'info_dict': {
  64             'id': '637842556329505',
  65             'ext': 'mp4',
  66             'title': 're:Did you know Kei Nishikori is the first Asian man to ever reach a Grand Slam',
  67             'uploader': 'Tennis on Facebook',
  68             'upload_date': '20140908',
  69             'timestamp': 1410199200,
  70         },
  71         'skip': 'Requires logging in',
  72     }, {
  73         'url': 'https://www.facebook.com/video.php?v=274175099429670',
  74         'info_dict': {
  75             'id': '274175099429670',
  76             'ext': 'mp4',
  77             'title': 'Asif Nawab Butt posted a video to his Timeline.',
  78             'uploader': 'Asif Nawab Butt',
  79             'upload_date': '20140506',
  80             'timestamp': 1399398998,
  81             'thumbnail': r're:^https?://.*',
  82         },
  83         'expected_warnings': [
  84             'title'
  85         ]
  86     }, {
  87         'note': 'Video with DASH manifest',
  88         'url': 'https://www.facebook.com/video.php?v=957955867617029',
  89         'md5': 'b2c28d528273b323abe5c6ab59f0f030',
  90         'info_dict': {
  91             'id': '957955867617029',
  92             'ext': 'mp4',
  93             'title': 'When you post epic content on instagram.com/433 8 million followers, this is ...',
  94             'uploader': 'Demy de Zeeuw',
  95             'upload_date': '20160110',
  96             'timestamp': 1452431627,
  97         },
  98         'skip': 'Requires logging in',
  99     }, {
 100         'url': 'https://www.facebook.com/maxlayn/posts/10153807558977570',
 101         'md5': '037b1fa7f3c2d02b7a0d7bc16031ecc6',
 102         'info_dict': {
 103             'id': '544765982287235',
 104             'ext': 'mp4',
 105             'title': '"What are you doing running in the snow?"',
 106             'uploader': 'FailArmy',
 107         },
 108         'skip': 'Video gone',
 109     }, {
 110         'url': 'https://m.facebook.com/story.php?story_fbid=1035862816472149&id=116132035111903',
 111         'md5': '1deb90b6ac27f7efcf6d747c8a27f5e3',
 112         'info_dict': {
 113             'id': '1035862816472149',
 114             'ext': 'mp4',
 115             'title': 'What the Flock Is Going On In New Zealand  Credit: ViralHog',
 116             'uploader': 'S. Saint',
 117         },
 118         'skip': 'Video gone',
 119     }, {
 120         'note': 'swf params escaped',
 121         'url': 'https://www.facebook.com/barackobama/posts/10153664894881749',
 122         'md5': '97ba073838964d12c70566e0085c2b91',
 123         'info_dict': {
 124             'id': '10153664894881749',
 125             'ext': 'mp4',
 126             'title': 'Average time to confirm recent Supreme Court nominees: 67 days Longest it\'s t...',
 127             'thumbnail': r're:^https?://.*',
 128             'timestamp': 1456259628,
 129             'upload_date': '20160223',
 130             'uploader': 'Barack Obama',
 131         },
 132     }, {
 133         # have 1080P, but only up to 720p in swf params
 134         'url': 'https://www.facebook.com/cnn/videos/10155529876156509/',
 135         'md5': '0d9813160b146b3bc8744e006027fcc6',
 136         'info_dict': {
 137             'id': '10155529876156509',
 138             'ext': 'mp4',
 139             'title': 'She survived the holocaust — and years later, she’s getting her citizenship s...',
 140             'timestamp': 1477818095,
 141             'upload_date': '20161030',
 142             'uploader': 'CNN',
 143             'thumbnail': r're:^https?://.*',
 144         },
 145     }, {
 146         # bigPipe.onPageletArrive ... onPageletArrive pagelet_group_mall
 147         'url': 'https://www.facebook.com/yaroslav.korpan/videos/1417995061575415/',
 148         'info_dict': {
 149             'id': '1417995061575415',
 150             'ext': 'mp4',
 151             'title': 'md5:a7b86ca673f51800cd54687b7f4012fe',
 152             'timestamp': 1486648217,
 153             'upload_date': '20170209',
 154             'uploader': 'Yaroslav Korpan',
 155         },
 156         'params': {
 157             'skip_download': True,
 158         },
 159     }, {
 160         'url': 'https://www.facebook.com/LaGuiaDelVaron/posts/1072691702860471',
 161         'info_dict': {
 162             'id': '1072691702860471',
 163             'ext': 'mp4',
 164             'title': 'md5:ae2d22a93fbb12dad20dc393a869739d',
 165             'timestamp': 1477305000,
 166             'upload_date': '20161024',
 167             'uploader': 'La Guía Del Varón',
 168             'thumbnail': r're:^https?://.*',
 169         },
 170         'params': {
 171             'skip_download': True,
 172         },
 173     }, {
 174         'url': 'https://www.facebook.com/groups/1024490957622648/permalink/1396382447100162/',
 175         'info_dict': {
 176             'id': '1396382447100162',
 177             'ext': 'mp4',
 178             'title': 'md5:e2d2700afdf84e121f5d0f999bad13a3',
 179             'timestamp': 1486035494,
 180             'upload_date': '20170202',
 181             'uploader': 'Elisabeth Ahtn',
 182         },
 183         'params': {
 184             'skip_download': True,
 185         },
 186     }, {
 187         'url': 'https://www.facebook.com/video.php?v=10204634152394104',
 188         'only_matching': True,
 189     }, {
 190         'url': 'https://www.facebook.com/amogood/videos/1618742068337349/?fref=nf',
 191         'only_matching': True,
 192     }, {
 193         'url': 'https://www.facebook.com/ChristyClarkForBC/videos/vb.22819070941/10153870694020942/?type=2&theater',
 194         'only_matching': True,
 195     }, {
 196         'url': 'facebook:544765982287235',
 197         'only_matching': True,
 198     }, {
 199         'url': 'https://www.facebook.com/groups/164828000315060/permalink/764967300301124/',
 200         'only_matching': True,
 201     }, {
 202         'url': 'https://zh-hk.facebook.com/peoplespower/videos/1135894589806027/',
 203         'only_matching': True,
 204     }, {
 205         'url': 'https://www.facebookcorewwwi.onion/video.php?v=274175099429670',
 206         'only_matching': True,
 207     }, {
 208         # no title
 209         'url': 'https://www.facebook.com/onlycleverentertainment/videos/1947995502095005/',
 210         'only_matching': True,
 211     }]
 212
 213     @staticmethod
 214     def _extract_urls(webpage):
 215         urls = []
 216         for mobj in re.finditer(
 217                 r'<iframe[^>]+?src=(["\'])(?P<url>https?://www\.facebook\.com/(?:video/embed|plugins/video\.php).+?)\1',
 218                 webpage):
 219             urls.append(mobj.group('url'))
 220         # Facebook API embed
 221         # see https://developers.facebook.com/docs/plugins/embedded-video-player
 222         for mobj in re.finditer(r'''(?x)<div[^>]+
 223                 class=(?P<q1>[\'"])[^\'"]*\bfb-(?:video|post)\b[^\'"]*(?P=q1)[^>]+
 224                 data-href=(?P<q2>[\'"])(?P<url>(?:https?:)?//(?:www\.)?facebook.com/.+?)(?P=q2)''', webpage):
 225             urls.append(mobj.group('url'))
 226         return urls
 227
 228     def _login(self):
 229         (useremail, password) = self._get_login_info()
 230         if useremail is None:
 231             return
 232
 233         login_page_req = sanitized_Request(self._LOGIN_URL)
 234         self._set_cookie('facebook.com', 'locale', 'en_US')
 235         login_page = self._download_webpage(login_page_req, None,
 236                                             note='Downloading login page',
 237                                             errnote='Unable to download login page')
 238         lsd = self._search_regex(
 239             r'<input type="hidden" name="lsd" value="([^"]*)"',
 240             login_page, 'lsd')
 241         lgnrnd = self._search_regex(r'name="lgnrnd" value="([^"]*?)"', login_page, 'lgnrnd')
 242
 243         login_form = {
 244             'email': useremail,
 245             'pass': password,
 246             'lsd': lsd,
 247             'lgnrnd': lgnrnd,
 248             'next': 'http://facebook.com/home.php',
 249             'default_persistent': '0',
 250             'legacy_return': '1',
 251             'timezone': '-60',
 252             'trynum': '1',
 253         }
 254         request = sanitized_Request(self._LOGIN_URL, urlencode_postdata(login_form))
 255         request.add_header('Content-Type', 'application/x-www-form-urlencoded')
 256         try:
 257             login_results = self._download_webpage(request, None,
 258                                                    note='Logging in', errnote='unable to fetch login page')
 259             if re.search(r'<form(.*)name="login"(.*)</form>', login_results) is not None:
 260                 error = self._html_search_regex(
 261                     r'(?s)<div[^>]+class=(["\']).*?login_error_box.*?\1[^>]*><div[^>]*>.*?</div><div[^>]*>(?P<error>.+?)</div>',
 262                     login_results, 'login error', default=None, group='error')
 263                 if error:
 264                     raise ExtractorError('Unable to login: %s' % error, expected=True)
 265                 self._downloader.report_warning('unable to log in: bad username/password, or exceeded login rate limit (~3/min). Check credentials or wait.')
 266                 return
 267
 268             fb_dtsg = self._search_regex(
 269                 r'name="fb_dtsg" value="(.+?)"', login_results, 'fb_dtsg', default=None)
 270             h = self._search_regex(
 271                 r'name="h"\s+(?:\w+="[^"]+"\s+)*?value="([^"]+)"', login_results, 'h', default=None)
 272
 273             if not fb_dtsg or not h:
 274                 return
 275
 276             check_form = {
 277                 'fb_dtsg': fb_dtsg,
 278                 'h': h,
 279                 'name_action_selected': 'dont_save',
 280             }
 281             check_req = sanitized_Request(self._CHECKPOINT_URL, urlencode_postdata(check_form))
 282             check_req.add_header('Content-Type', 'application/x-www-form-urlencoded')
 283             check_response = self._download_webpage(check_req, None,
 284                                                     note='Confirming login')
 285             if re.search(r'id="checkpointSubmitButton"', check_response) is not None:
 286                 self._downloader.report_warning('Unable to confirm login, you have to login in your browser and authorize the login.')
 287         except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
 288             self._downloader.report_warning('unable to log in: %s' % error_to_compat_str(err))
 289             return
 290
 291     def _real_initialize(self):
 292         self._login()
 293
 294     def _extract_from_url(self, url, video_id, fatal_if_no_video=True):
 295         req = sanitized_Request(url)
 296         req.add_header('User-Agent', self._CHROME_USER_AGENT)
 297         webpage = self._download_webpage(req, video_id)
 298
 299         video_data = None
 300
 301         def extract_video_data(instances):
 302             for item in instances:
 303                 if item[1][0] == 'VideoConfig':
 304                     video_item = item[2][0]
 305                     if video_item.get('video_id'):
 306                         return video_item['videoData']
 307
 308         server_js_data = self._parse_json(self._search_regex(
 309             r'handleServerJS\(({.+})(?:\);|,")', webpage,
 310             'server js data', default='{}'), video_id, fatal=False)
 311
 312         if server_js_data:
 313             video_data = extract_video_data(server_js_data.get('instances', []))
 314
 315         if not video_data:
 316             server_js_data = self._parse_json(
 317                 self._search_regex(
 318                     r'bigPipe\.onPageletArrive\(({.+?})\)\s*;\s*}\s*\)\s*,\s*["\']onPageletArrive\s+(?:stream_pagelet|pagelet_group_mall|permalink_video_pagelet)',
 319                     webpage, 'js data', default='{}'),
 320                 video_id, transform_source=js_to_json, fatal=False)
 321             if server_js_data:
 322                 video_data = extract_video_data(try_get(
 323                     server_js_data, lambda x: x['jsmods']['instances'],
 324                     list) or [])
 325
 326         if not video_data:
 327             if not fatal_if_no_video:
 328                 return webpage, False
 329             m_msg = re.search(r'class="[^"]*uiInterstitialContent[^"]*"><div>(.*?)</div>', webpage)
 330             if m_msg is not None:
 331                 raise ExtractorError(
 332                     'The video is not available, Facebook said: "%s"' % m_msg.group(1),
 333                     expected=True)
 334             elif '>You must log in to continue' in webpage:
 335                 self.raise_login_required()
 336             else:
 337                 raise ExtractorError('Cannot parse data')
 338
 339         formats = []
 340         for f in video_data:
 341             format_id = f['stream_type']
 342             if f and isinstance(f, dict):
 343                 f = [f]
 344             if not f or not isinstance(f, list):
 345                 continue
 346             for quality in ('sd', 'hd'):
 347                 for src_type in ('src', 'src_no_ratelimit'):
 348                     src = f[0].get('%s_%s' % (quality, src_type))
 349                     if src:
 350                         preference = -10 if format_id == 'progressive' else 0
 351                         if quality == 'hd':
 352                             preference += 5
 353                         formats.append({
 354                             'format_id': '%s_%s_%s' % (format_id, quality, src_type),
 355                             'url': src,
 356                             'preference': preference,
 357                         })
 358             dash_manifest = f[0].get('dash_manifest')
 359             if dash_manifest:
 360                 formats.extend(self._parse_mpd_formats(
 361                     compat_etree_fromstring(compat_urllib_parse_unquote_plus(dash_manifest))))
 362         if not formats:
 363             raise ExtractorError('Cannot find video formats')
 364
 365         self._sort_formats(formats)
 366
 367         video_title = self._html_search_regex(
 368             r'<h2\s+[^>]*class="uiHeaderTitle"[^>]*>([^<]*)</h2>', webpage,
 369             'title', default=None)
 370         if not video_title:
 371             video_title = self._html_search_regex(
 372                 r'(?s)<span class="fbPhotosPhotoCaption".*?id="fbPhotoPageCaption"><span class="hasCaption">(.*?)</span>',
 373                 webpage, 'alternative title', default=None)
 374         if not video_title:
 375             video_title = self._html_search_meta(
 376                 'description', webpage, 'title', default=None)
 377         if video_title:
 378             video_title = limit_length(video_title, 80)
 379         else:
 380             video_title = 'Facebook video #%s' % video_id
 381         uploader = clean_html(get_element_by_id(
 382             'fbPhotoPageAuthorName', webpage)) or self._search_regex(
 383             r'ownerName\s*:\s*"([^"]+)"', webpage, 'uploader', fatal=False)
 384         timestamp = int_or_none(self._search_regex(
 385             r'<abbr[^>]+data-utime=["\'](\d+)', webpage,
 386             'timestamp', default=None))
 387         thumbnail = self._og_search_thumbnail(webpage)
 388
 389         info_dict = {
 390             'id': video_id,
 391             'title': video_title,
 392             'formats': formats,
 393             'uploader': uploader,
 394             'timestamp': timestamp,
 395             'thumbnail': thumbnail,
 396         }
 397
 398         return webpage, info_dict
 399
 400     def _real_extract(self, url):
 401         video_id = self._match_id(url)
 402
 403         real_url = self._VIDEO_PAGE_TEMPLATE % video_id if url.startswith('facebook:') else url
 404         webpage, info_dict = self._extract_from_url(real_url, video_id, fatal_if_no_video=False)
 405
 406         if info_dict:
 407             return info_dict
 408
 409         if '/posts/' in url:
 410             entries = [
 411                 self.url_result('facebook:%s' % vid, FacebookIE.ie_key())
 412                 for vid in self._parse_json(
 413                     self._search_regex(
 414                         r'(["\'])video_ids\1\s*:\s*(?P<ids>\[.+?\])',
 415                         webpage, 'video ids', group='ids'),
 416                     video_id)]
 417
 418             return self.playlist_result(entries, video_id)
 419         else:
 420             _, info_dict = self._extract_from_url(
 421                 self._VIDEO_PAGE_TEMPLATE % video_id,
 422                 video_id, fatal_if_no_video=True)
 423             return info_dict
 424
 425
 426 class FacebookPluginsVideoIE(InfoExtractor):
 427     _VALID_URL = r'https?://(?:[\w-]+\.)?facebook\.com/plugins/video\.php\?.*?\bhref=(?P<id>https.+)'
 428
 429     _TESTS = [{
 430         'url': 'https://www.facebook.com/plugins/video.php?href=https%3A%2F%2Fwww.facebook.com%2Fgov.sg%2Fvideos%2F10154383743583686%2F&show_text=0&width=560',
 431         'md5': '5954e92cdfe51fe5782ae9bda7058a07',
 432         'info_dict': {
 433             'id': '10154383743583686',
 434             'ext': 'mp4',
 435             'title': 'What to do during the haze?',
 436             'uploader': 'Gov.sg',
 437             'upload_date': '20160826',
 438             'timestamp': 1472184808,
 439         },
 440         'add_ie': [FacebookIE.ie_key()],
 441     }, {
 442         'url': 'https://www.facebook.com/plugins/video.php?href=https%3A%2F%2Fwww.facebook.com%2Fvideo.php%3Fv%3D10204634152394104',
 443         'only_matching': True,
 444     }, {
 445         'url': 'https://www.facebook.com/plugins/video.php?href=https://www.facebook.com/gov.sg/videos/10154383743583686/&show_text=0&width=560',
 446         'only_matching': True,
 447     }]
 448
 449     def _real_extract(self, url):
 450         return self.url_result(
 451             compat_urllib_parse_unquote(self._match_id(url)),
 452             FacebookIE.ie_key())