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_http_client,
  10     compat_str,
  11     compat_urllib_error,
  12     compat_urllib_parse_unquote,
  13     compat_urllib_request,
  14 )
  15 from ..utils import (
  16     ExtractorError,
  17     int_or_none,
  18     limit_length,
  19     urlencode_postdata,
  20 )
  21
  22
  23 class FacebookIE(InfoExtractor):
  24     _VALID_URL = r'''(?x)
  25         https?://(?:\w+\.)?facebook\.com/
  26         (?:[^#]*?\#!/)?
  27         (?:
  28             (?:video/video\.php|photo\.php|video\.php|video/embed)\?(?:.*?)
  29             (?:v|video_id)=|
  30             [^/]+/videos/(?:[^/]+/)?
  31         )
  32         (?P<id>[0-9]+)
  33         (?:.*)'''
  34     _LOGIN_URL = 'https://www.facebook.com/login.php?next=http%3A%2F%2Ffacebook.com%2Fhome.php&login_attempt=1'
  35     _CHECKPOINT_URL = 'https://www.facebook.com/checkpoint/?next=http%3A%2F%2Ffacebook.com%2Fhome.php&_fb_noscript=1'
  36     _NETRC_MACHINE = 'facebook'
  37     IE_NAME = 'facebook'
  38     _TESTS = [{
  39         'url': 'https://www.facebook.com/video.php?v=637842556329505&fref=nf',
  40         'md5': '6a40d33c0eccbb1af76cf0485a052659',
  41         'info_dict': {
  42             'id': '637842556329505',
  43             'ext': 'mp4',
  44             'title': 're:Did you know Kei Nishikori is the first Asian man to ever reach a Grand Slam',
  45         }
  46     }, {
  47         'note': 'Video without discernible title',
  48         'url': 'https://www.facebook.com/video.php?v=274175099429670',
  49         'info_dict': {
  50             'id': '274175099429670',
  51             'ext': 'mp4',
  52             'title': 'Facebook video #274175099429670',
  53         },
  54         'expected_warnings': [
  55             'title'
  56         ]
  57     }, {
  58         'url': 'https://www.facebook.com/video.php?v=10204634152394104',
  59         'only_matching': True,
  60     }, {
  61         'url': 'https://www.facebook.com/amogood/videos/1618742068337349/?fref=nf',
  62         'only_matching': True,
  63     }, {
  64         'url': 'https://www.facebook.com/ChristyClarkForBC/videos/vb.22819070941/10153870694020942/?type=2&theater',
  65         'only_matching': True,
  66     }]
  67
  68     def _login(self):
  69         (useremail, password) = self._get_login_info()
  70         if useremail is None:
  71             return
  72
  73         login_page_req = compat_urllib_request.Request(self._LOGIN_URL)
  74         login_page_req.add_header('Cookie', 'locale=en_US')
  75         login_page = self._download_webpage(login_page_req, None,
  76                                             note='Downloading login page',
  77                                             errnote='Unable to download login page')
  78         lsd = self._search_regex(
  79             r'<input type="hidden" name="lsd" value="([^"]*)"',
  80             login_page, 'lsd')
  81         lgnrnd = self._search_regex(r'name="lgnrnd" value="([^"]*?)"', login_page, 'lgnrnd')
  82
  83         login_form = {
  84             'email': useremail,
  85             'pass': password,
  86             'lsd': lsd,
  87             'lgnrnd': lgnrnd,
  88             'next': 'http://facebook.com/home.php',
  89             'default_persistent': '0',
  90             'legacy_return': '1',
  91             'timezone': '-60',
  92             'trynum': '1',
  93         }
  94         request = compat_urllib_request.Request(self._LOGIN_URL, urlencode_postdata(login_form))
  95         request.add_header('Content-Type', 'application/x-www-form-urlencoded')
  96         try:
  97             login_results = self._download_webpage(request, None,
  98                                                    note='Logging in', errnote='unable to fetch login page')
  99             if re.search(r'<form(.*)name="login"(.*)</form>', login_results) is not None:
 100                 self._downloader.report_warning('unable to log in: bad username/password, or exceded login rate limit (~3/min). Check credentials or wait.')
 101                 return
 102
 103             check_form = {
 104                 'fb_dtsg': self._search_regex(r'name="fb_dtsg" value="(.+?)"', login_results, 'fb_dtsg'),
 105                 'h': self._search_regex(
 106                     r'name="h"\s+(?:\w+="[^"]+"\s+)*?value="([^"]+)"', login_results, 'h'),
 107                 'name_action_selected': 'dont_save',
 108             }
 109             check_req = compat_urllib_request.Request(self._CHECKPOINT_URL, urlencode_postdata(check_form))
 110             check_req.add_header('Content-Type', 'application/x-www-form-urlencoded')
 111             check_response = self._download_webpage(check_req, None,
 112                                                     note='Confirming login')
 113             if re.search(r'id="checkpointSubmitButton"', check_response) is not None:
 114                 self._downloader.report_warning('Unable to confirm login, you have to login in your brower and authorize the login.')
 115         except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
 116             self._downloader.report_warning('unable to log in: %s' % compat_str(err))
 117             return
 118
 119     def _real_initialize(self):
 120         self._login()
 121
 122     def _real_extract(self, url):
 123         video_id = self._match_id(url)
 124         url = 'https://www.facebook.com/video/video.php?v=%s' % video_id
 125         webpage = self._download_webpage(url, video_id)
 126
 127         BEFORE = '{swf.addParam(param[0], param[1]);});\n'
 128         AFTER = '.forEach(function(variable) {swf.addVariable(variable[0], variable[1]);});'
 129         m = re.search(re.escape(BEFORE) + '(.*?)' + re.escape(AFTER), webpage)
 130         if not m:
 131             m_msg = re.search(r'class="[^"]*uiInterstitialContent[^"]*"><div>(.*?)</div>', webpage)
 132             if m_msg is not None:
 133                 raise ExtractorError(
 134                     'The video is not available, Facebook said: "%s"' % m_msg.group(1),
 135                     expected=True)
 136             else:
 137                 raise ExtractorError('Cannot parse data')
 138         data = dict(json.loads(m.group(1)))
 139         params_raw = compat_urllib_parse_unquote(data['params'])
 140         params = json.loads(params_raw)
 141         video_data = params['video_data'][0]
 142
 143         formats = []
 144         for quality in ['sd', 'hd']:
 145             src = video_data.get('%s_src' % quality)
 146             if src is not None:
 147                 formats.append({
 148                     'format_id': quality,
 149                     'url': src,
 150                 })
 151         if not formats:
 152             raise ExtractorError('Cannot find video formats')
 153
 154         video_title = self._html_search_regex(
 155             r'<h2\s+[^>]*class="uiHeaderTitle"[^>]*>([^<]*)</h2>', webpage, 'title',
 156             default=None)
 157         if not video_title:
 158             video_title = self._html_search_regex(
 159                 r'(?s)<span class="fbPhotosPhotoCaption".*?id="fbPhotoPageCaption"><span class="hasCaption">(.*?)</span>',
 160                 webpage, 'alternative title', fatal=False)
 161             video_title = limit_length(video_title, 80)
 162         if not video_title:
 163             video_title = 'Facebook video #%s' % video_id
 164
 165         return {
 166             'id': video_id,
 167             'title': video_title,
 168             'formats': formats,
 169             'duration': int_or_none(video_data.get('video_duration')),
 170             'thumbnail': video_data.get('thumbnail_src'),
 171         }