yt_dlp/extractor/viously.py

   1 import base64
   2 import re
   3
   4 from .common import InfoExtractor
   5 from ..utils import (
   6     extract_attributes,
   7     int_or_none,
   8     parse_iso8601,
   9 )
  10 from ..utils.traversal import traverse_obj
  11
  12
  13 class ViouslyIE(InfoExtractor):
  14     _VALID_URL = False
  15     _WEBPAGE_TESTS = [{
  16         'url': 'http://www.turbo.fr/videos-voiture/454443-turbo-du-07-09-2014-renault-twingo-3-bentley-continental-gt-speed-ces-guide-achat-dacia.html',
  17         'md5': '37a6c3381599381ff53a7e1e0575c0bc',
  18         'info_dict': {
  19             'id': 'F_xQzS2jwb3',
  20             'ext': 'mp4',
  21             'title': 'Turbo du 07/09/2014\xa0: Renault Twingo 3, Bentley Continental GT Speed, CES, Guide Achat Dacia...',
  22             'description': 'Turbo du 07/09/2014\xa0: Renault Twingo 3, Bentley Continental GT Speed, CES, Guide Achat Dacia...',
  23             'age_limit': 0,
  24             'upload_date': '20230328',
  25             'timestamp': 1680037507,
  26             'duration': 3716,
  27             'categories': ['motors'],
  28         }
  29     }]
  30
  31     def _extract_from_webpage(self, url, webpage):
  32         viously_players = re.findall(r'<div[^>]*class="(?:[^"]*\s)?v(?:iou)?sly-player(?:\s[^"]*)?"[^>]*>', webpage)
  33         if not viously_players:
  34             return
  35
  36         def custom_decode(text):
  37             STANDARD_ALPHABET = 'ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/='
  38             CUSTOM_ALPHABET = 'VIOUSLYABCDEFGHJKMNPQRTWXZviouslyabcdefghjkmnpqrtwxz9876543210+/='
  39             data = base64.b64decode(text.translate(str.maketrans(CUSTOM_ALPHABET, STANDARD_ALPHABET)))
  40             return data.decode('utf-8').strip('\x00')
  41
  42         for video_id in traverse_obj(viously_players, (..., {extract_attributes}, 'id')):
  43             formats = self._extract_m3u8_formats(
  44                 f'https://www.viously.com/video/hls/{video_id}/index.m3u8', video_id, fatal=False)
  45             if not formats:
  46                 continue
  47             data = self._download_json(
  48                 f'https://www.viously.com/export/json/{video_id}', video_id,
  49                 transform_source=custom_decode, fatal=False)
  50             yield {
  51                 'id': video_id,
  52                 'formats': formats,
  53                 **traverse_obj(data, ('video', {
  54                     'title': ('title', {str}),
  55                     'description': ('description', {str}),
  56                     'duration': ('duration', {int_or_none}),
  57                     'timestamp': ('iso_date', {parse_iso8601}),
  58                     'categories': ('category', 'name', {str}, {lambda x: [x] if x else None}),
  59                 })),
  60             }