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

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import json
   5 import re
   6
   7 from .common import InfoExtractor
   8 from ..utils import (
   9     ExtractorError,
  10     parse_iso8601,
  11     sanitized_Request,
  12 )
  13
  14
  15 class VesselIE(InfoExtractor):
  16     _VALID_URL = r'https?://(?:www\.)?vessel\.com/(?:videos|embed)/(?P<id>[0-9a-zA-Z]+)'
  17     _API_URL_TEMPLATE = 'https://www.vessel.com/api/view/items/%s'
  18     _LOGIN_URL = 'https://www.vessel.com/api/account/login'
  19     _NETRC_MACHINE = 'vessel'
  20     _TESTS = [{
  21         'url': 'https://www.vessel.com/videos/HDN7G5UMs',
  22         'md5': '455cdf8beb71c6dd797fd2f3818d05c4',
  23         'info_dict': {
  24             'id': 'HDN7G5UMs',
  25             'ext': 'mp4',
  26             'title': 'Nvidia GeForce GTX Titan X - The Best Video Card on the Market?',
  27             'thumbnail': 're:^https?://.*\.jpg$',
  28             'upload_date': '20150317',
  29             'description': 'Did Nvidia pull out all the stops on the Titan X, or does its performance leave something to be desired?',
  30             'timestamp': int,
  31         },
  32     }, {
  33         'url': 'https://www.vessel.com/embed/G4U7gUJ6a?w=615&h=346',
  34         'only_matching': True,
  35     }]
  36
  37     @staticmethod
  38     def _extract_urls(webpage):
  39         return [url for _, url in re.findall(
  40             r'<iframe[^>]+src=(["\'])((?:https?:)?//(?:www\.)?vessel\.com/embed/[0-9a-zA-Z]+.*?)\1',
  41             webpage)]
  42
  43     @staticmethod
  44     def make_json_request(url, data):
  45         payload = json.dumps(data).encode('utf-8')
  46         req = sanitized_Request(url, payload)
  47         req.add_header('Content-Type', 'application/json; charset=utf-8')
  48         return req
  49
  50     @staticmethod
  51     def find_assets(data, asset_type, asset_id=None):
  52         for asset in data.get('assets', []):
  53             if not asset.get('type') == asset_type:
  54                 continue
  55             elif asset_id is not None and not asset.get('id') == asset_id:
  56                 continue
  57             else:
  58                 yield asset
  59
  60     def _check_access_rights(self, data):
  61         access_info = data.get('__view', {})
  62         if not access_info.get('allow_access', True):
  63             err_code = access_info.get('error_code') or ''
  64             if err_code == 'ITEM_PAID_ONLY':
  65                 raise ExtractorError(
  66                     'This video requires subscription.', expected=True)
  67             else:
  68                 raise ExtractorError(
  69                     'Access to this content is restricted. (%s said: %s)' % (self.IE_NAME, err_code), expected=True)
  70
  71     def _login(self):
  72         (username, password) = self._get_login_info()
  73         if username is None:
  74             return
  75         self.report_login()
  76         data = {
  77             'client_id': 'web',
  78             'type': 'password',
  79             'user_key': username,
  80             'password': password,
  81         }
  82         login_request = VesselIE.make_json_request(self._LOGIN_URL, data)
  83         self._download_webpage(login_request, None, False, 'Wrong login info')
  84
  85     def _real_initialize(self):
  86         self._login()
  87
  88     def _real_extract(self, url):
  89         video_id = self._match_id(url)
  90
  91         webpage = self._download_webpage(url, video_id)
  92         data = self._parse_json(self._search_regex(
  93             r'App\.bootstrapData\((.*?)\);', webpage, 'data'), video_id)
  94         asset_id = data['model']['data']['id']
  95
  96         req = VesselIE.make_json_request(
  97             self._API_URL_TEMPLATE % asset_id, {'client': 'web'})
  98         data = self._download_json(req, video_id)
  99         video_asset_id = data.get('main_video_asset')
 100
 101         self._check_access_rights(data)
 102
 103         try:
 104             video_asset = next(
 105                 VesselIE.find_assets(data, 'video', asset_id=video_asset_id))
 106         except StopIteration:
 107             raise ExtractorError('No video assets found')
 108
 109         formats = []
 110         for f in video_asset.get('sources', []):
 111             location = f.get('location')
 112             if not location:
 113                 continue
 114             name = f.get('name')
 115             if name == 'hls-index':
 116                 formats.extend(self._extract_m3u8_formats(
 117                     location, video_id, ext='mp4',
 118                     entry_protocol='m3u8_native', m3u8_id='m3u8', fatal=False))
 119             elif name == 'dash-index':
 120                 formats.extend(self._extract_mpd_formats(
 121                     location, video_id, mpd_id='dash', fatal=False))
 122             else:
 123                 formats.append({
 124                     'format_id': name,
 125                     'tbr': f.get('bitrate'),
 126                     'height': f.get('height'),
 127                     'width': f.get('width'),
 128                     'url': location,
 129                 })
 130         self._sort_formats(formats)
 131
 132         thumbnails = []
 133         for im_asset in VesselIE.find_assets(data, 'image'):
 134             thumbnails.append({
 135                 'url': im_asset['location'],
 136                 'width': im_asset.get('width', 0),
 137                 'height': im_asset.get('height', 0),
 138             })
 139
 140         return {
 141             'id': video_id,
 142             'title': data['title'],
 143             'formats': formats,
 144             'thumbnails': thumbnails,
 145             'description': data.get('short_description'),
 146             'duration': data.get('duration'),
 147             'comment_count': data.get('comment_count'),
 148             'like_count': data.get('like_count'),
 149             'view_count': data.get('view_count'),
 150             'timestamp': parse_iso8601(data.get('released_at')),
 151         }