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