_ Git - youtube-dl/blob - youtube_dl/extractor/facebook.py

   1 from __future__ import unicode_literals
   2
   3 import json
   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     error_to_compat_str,
  17     ExtractorError,
  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                         )|
  42                     facebook:
  43                 )
  44                 (?P<id>[0-9]+)
  45                 '''
  46     _LOGIN_URL = 'https://www.facebook.com/login.php?next=http%3A%2F%2Ffacebook.com%2Fhome.php&login_attempt=1'
  47     _CHECKPOINT_URL = 'https://www.facebook.com/checkpoint/?next=http%3A%2F%2Ffacebook.com%2Fhome.php&_fb_noscript=1'
  48     _NETRC_MACHINE = 'facebook'
  49     IE_NAME = 'facebook'
  50
  51     _CHROME_USER_AGENT = 'Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/48.0.2564.97 Safari/537.36'
  52
  53     _TESTS = [{
  54         'url': 'https://www.facebook.com/video.php?v=637842556329505&fref=nf',
  55         'md5': '6a40d33c0eccbb1af76cf0485a052659',
  56         'info_dict': {
  57             'id': '637842556329505',
  58             'ext': 'mp4',
  59             'title': 're:Did you know Kei Nishikori is the first Asian man to ever reach a Grand Slam',
  60             'uploader': 'Tennis on Facebook',
  61         }
  62     }, {
  63         'note': 'Video without discernible title',
  64         'url': 'https://www.facebook.com/video.php?v=274175099429670',
  65         'info_dict': {
  66             'id': '274175099429670',
  67             'ext': 'mp4',
  68             'title': 'Facebook video #274175099429670',
  69             'uploader': 'Asif Nawab Butt',
  70         },
  71         'expected_warnings': [
  72             'title'
  73         ]
  74     }, {
  75         'note': 'Video with DASH manifest',
  76         'url': 'https://www.facebook.com/video.php?v=957955867617029',
  77         'md5': '54706e4db4f5ad58fbad82dde1f1213f',
  78         'info_dict': {
  79             'id': '957955867617029',
  80             'ext': 'mp4',
  81             'title': 'When you post epic content on instagram.com/433 8 million followers, this is ...',
  82             'uploader': 'Demy de Zeeuw',
  83         },
  84     }, {
  85         'url': 'https://www.facebook.com/video.php?v=10204634152394104',
  86         'only_matching': True,
  87     }, {
  88         'url': 'https://www.facebook.com/amogood/videos/1618742068337349/?fref=nf',
  89         'only_matching': True,
  90     }, {
  91         'url': 'https://www.facebook.com/ChristyClarkForBC/videos/vb.22819070941/10153870694020942/?type=2&theater',
  92         'only_matching': True,
  93     }, {
  94         'url': 'facebook:544765982287235',
  95         'only_matching': True,
  96     }, {
  97         'url': 'https://m.facebook.com/story.php?story_fbid=1035862816472149&id=116132035111903',
  98         'only_matching': True,
  99     }]
 100
 101     def _login(self):
 102         (useremail, password) = self._get_login_info()
 103         if useremail is None:
 104             return
 105
 106         login_page_req = sanitized_Request(self._LOGIN_URL)
 107         self._set_cookie('facebook.com', 'locale', 'en_US')
 108         login_page = self._download_webpage(login_page_req, None,
 109                                             note='Downloading login page',
 110                                             errnote='Unable to download login page')
 111         lsd = self._search_regex(
 112             r'<input type="hidden" name="lsd" value="([^"]*)"',
 113             login_page, 'lsd')
 114         lgnrnd = self._search_regex(r'name="lgnrnd" value="([^"]*?)"', login_page, 'lgnrnd')
 115
 116         login_form = {
 117             'email': useremail,
 118             'pass': password,
 119             'lsd': lsd,
 120             'lgnrnd': lgnrnd,
 121             'next': 'http://facebook.com/home.php',
 122             'default_persistent': '0',
 123             'legacy_return': '1',
 124             'timezone': '-60',
 125             'trynum': '1',
 126         }
 127         request = sanitized_Request(self._LOGIN_URL, urlencode_postdata(login_form))
 128         request.add_header('Content-Type', 'application/x-www-form-urlencoded')
 129         try:
 130             login_results = self._download_webpage(request, None,
 131                                                    note='Logging in', errnote='unable to fetch login page')
 132             if re.search(r'<form(.*)name="login"(.*)</form>', login_results) is not None:
 133                 error = self._html_search_regex(
 134                     r'(?s)<div[^>]+class=(["\']).*?login_error_box.*?\1[^>]*><div[^>]*>.*?</div><div[^>]*>(?P<error>.+?)</div>',
 135                     login_results, 'login error', default=None, group='error')
 136                 if error:
 137                     raise ExtractorError('Unable to login: %s' % error, expected=True)
 138                 self._downloader.report_warning('unable to log in: bad username/password, or exceeded login rate limit (~3/min). Check credentials or wait.')
 139                 return
 140
 141             fb_dtsg = self._search_regex(
 142                 r'name="fb_dtsg" value="(.+?)"', login_results, 'fb_dtsg', default=None)
 143             h = self._search_regex(
 144                 r'name="h"\s+(?:\w+="[^"]+"\s+)*?value="([^"]+)"', login_results, 'h', default=None)
 145
 146             if not fb_dtsg or not h:
 147                 return
 148
 149             check_form = {
 150                 'fb_dtsg': fb_dtsg,
 151                 'h': h,
 152                 'name_action_selected': 'dont_save',
 153             }
 154             check_req = sanitized_Request(self._CHECKPOINT_URL, urlencode_postdata(check_form))
 155             check_req.add_header('Content-Type', 'application/x-www-form-urlencoded')
 156             check_response = self._download_webpage(check_req, None,
 157                                                     note='Confirming login')
 158             if re.search(r'id="checkpointSubmitButton"', check_response) is not None:
 159                 self._downloader.report_warning('Unable to confirm login, you have to login in your browser and authorize the login.')
 160         except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
 161             self._downloader.report_warning('unable to log in: %s' % error_to_compat_str(err))
 162             return
 163
 164     def _real_initialize(self):
 165         self._login()
 166
 167     def _real_extract(self, url):
 168         video_id = self._match_id(url)
 169         req = sanitized_Request('https://www.facebook.com/video/video.php?v=%s' % video_id)
 170         req.add_header('User-Agent', self._CHROME_USER_AGENT)
 171         webpage = self._download_webpage(req, video_id)
 172
 173         video_data = None
 174
 175         BEFORE = '{swf.addParam(param[0], param[1]);});\n'
 176         AFTER = '.forEach(function(variable) {swf.addVariable(variable[0], variable[1]);});'
 177         m = re.search(re.escape(BEFORE) + '(.*?)' + re.escape(AFTER), webpage)
 178         if m:
 179             data = dict(json.loads(m.group(1)))
 180             params_raw = compat_urllib_parse_unquote(data['params'])
 181             video_data = json.loads(params_raw)['video_data']
 182
 183         def video_data_list2dict(video_data):
 184             ret = {}
 185             for item in video_data:
 186                 format_id = item['stream_type']
 187                 ret.setdefault(format_id, []).append(item)
 188             return ret
 189
 190         if not video_data:
 191             server_js_data = self._parse_json(self._search_regex(
 192                 r'handleServerJS\(({.+})\);', webpage, 'server js data'), video_id)
 193             for item in server_js_data.get('instances', []):
 194                 if item[1][0] == 'VideoConfig':
 195                     video_data = video_data_list2dict(item[2][0]['videoData'])
 196                     break
 197
 198         if not video_data:
 199             m_msg = re.search(r'class="[^"]*uiInterstitialContent[^"]*"><div>(.*?)</div>', webpage)
 200             if m_msg is not None:
 201                 raise ExtractorError(
 202                     'The video is not available, Facebook said: "%s"' % m_msg.group(1),
 203                     expected=True)
 204             else:
 205                 raise ExtractorError('Cannot parse data')
 206
 207         formats = []
 208         for format_id, f in video_data.items():
 209             if not f or not isinstance(f, list):
 210                 continue
 211             for quality in ('sd', 'hd'):
 212                 for src_type in ('src', 'src_no_ratelimit'):
 213                     src = f[0].get('%s_%s' % (quality, src_type))
 214                     if src:
 215                         formats.append({
 216                             'format_id': '%s_%s_%s' % (format_id, quality, src_type),
 217                             'url': src,
 218                             'preference': -10 if format_id == 'progressive' else 0,
 219                         })
 220             dash_manifest = f[0].get('dash_manifest')
 221             if dash_manifest:
 222                 formats.extend(self._parse_mpd_formats(
 223                     compat_etree_fromstring(compat_urllib_parse_unquote_plus(dash_manifest))))
 224         if not formats:
 225             raise ExtractorError('Cannot find video formats')
 226
 227         self._sort_formats(formats)
 228
 229         video_title = self._html_search_regex(
 230             r'<h2\s+[^>]*class="uiHeaderTitle"[^>]*>([^<]*)</h2>', webpage, 'title',
 231             default=None)
 232         if not video_title:
 233             video_title = self._html_search_regex(
 234                 r'(?s)<span class="fbPhotosPhotoCaption".*?id="fbPhotoPageCaption"><span class="hasCaption">(.*?)</span>',
 235                 webpage, 'alternative title', default=None)
 236             video_title = limit_length(video_title, 80)
 237         if not video_title:
 238             video_title = 'Facebook video #%s' % video_id
 239         uploader = clean_html(get_element_by_id('fbPhotoPageAuthorName', webpage))
 240
 241         return {
 242             'id': video_id,
 243             'title': video_title,
 244             'formats': formats,
 245             'uploader': uploader,
 246         }
 247
 248
 249 class FacebookPostIE(InfoExtractor):
 250     IE_NAME = 'facebook:post'
 251     _VALID_URL = r'https?://(?:\w+\.)?facebook\.com/[^/]+/posts/(?P<id>\d+)'
 252     _TEST = {
 253         'url': 'https://www.facebook.com/maxlayn/posts/10153807558977570',
 254         'md5': '037b1fa7f3c2d02b7a0d7bc16031ecc6',
 255         'info_dict': {
 256             'id': '544765982287235',
 257             'ext': 'mp4',
 258             'title': '"What are you doing running in the snow?"',
 259             'uploader': 'FailArmy',
 260         }
 261     }
 262
 263     def _real_extract(self, url):
 264         post_id = self._match_id(url)
 265
 266         webpage = self._download_webpage(url, post_id)
 267
 268         entries = [
 269             self.url_result('facebook:%s' % video_id, FacebookIE.ie_key())
 270             for video_id in self._parse_json(
 271                 self._search_regex(
 272                     r'(["\'])video_ids\1\s*:\s*(?P<ids>\[.+?\])',
 273                     webpage, 'video ids', group='ids'),
 274                 post_id)]
 275
 276         return self.playlist_result(entries, post_id)