youtube_dl/extractor/facebook.py

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