yt_dlp/extractor/jstream.py

   1 import base64
   2 import json
   3 import re
   4
   5 from .common import InfoExtractor
   6 from ..utils import (
   7     float_or_none,
   8     js_to_json,
   9     remove_start,
  10 )
  11
  12
  13 class JStreamIE(InfoExtractor):
  14     # group "id" only exists for compliance, not directly used in requests
  15     # also all components are mandatory
  16     _VALID_URL = r'jstream:(?P<host>www\d+):(?P<id>(?P<publisher>[a-z0-9]+):(?P<mid>\d+))'
  17
  18     _TESTS = [{
  19         'url': 'jstream:www50:eqd638pvwx:752',
  20         'info_dict': {
  21             'id': 'eqd638pvwx:752',
  22             'ext': 'mp4',
  23             'title': '阪神淡路大震災 激震の記録2020年版　解説動画',
  24             'duration': 672,
  25             'thumbnail': r're:https?://eqd638pvwx\.eq\.webcdn\.stream\.ne\.jp/.+\.jpg',
  26         },
  27     }]
  28
  29     def _parse_jsonp(self, callback, string, video_id):
  30         return self._search_json(rf'\s*{re.escape(callback)}\s*\(', string, callback, video_id)
  31
  32     def _find_formats(self, video_id, movie_list_hls, host, publisher, subtitles):
  33         for value in movie_list_hls:
  34             text = value.get('text') or ''
  35             if not text.startswith('auto'):
  36                 continue
  37             m3u8_id = remove_start(remove_start(text, 'auto'), '_') or None
  38             fmts, subs = self._extract_m3u8_formats_and_subtitles(
  39                 f'https://{publisher}.eq.webcdn.stream.ne.jp/{host}/{publisher}/jmc_pub/{value.get("url")}', video_id, 'mp4', m3u8_id=m3u8_id)
  40             self._merge_subtitles(subs, target=subtitles)
  41             yield from fmts
  42
  43     def _real_extract(self, url):
  44         host, publisher, mid, video_id = self._match_valid_url(url).group('host', 'publisher', 'mid', 'id')
  45         video_info_jsonp = self._download_webpage(
  46             f'https://{publisher}.eq.webcdn.stream.ne.jp/{host}/{publisher}/jmc_pub/eq_meta/v1/{mid}.jsonp',
  47             video_id, 'Requesting video info')
  48         video_info = self._parse_jsonp('metaDataResult', video_info_jsonp, video_id)['movie']
  49         subtitles = {}
  50         formats = list(self._find_formats(video_id, video_info.get('movie_list_hls'), host, publisher, subtitles))
  51         self._remove_duplicate_formats(formats)
  52         return {
  53             'id': video_id,
  54             'title': video_info.get('title'),
  55             'duration': float_or_none(video_info.get('duration')),
  56             'thumbnail': video_info.get('thumbnail_url'),
  57             'formats': formats,
  58             'subtitles': subtitles,
  59         }
  60
  61     @classmethod
  62     def _extract_embed_urls(cls, url, webpage):
  63         # check for eligiblity of webpage
  64         # https://support.eq.stream.co.jp/hc/ja/articles/115008388147-%E3%83%97%E3%83%AC%E3%82%A4%E3%83%A4%E3%83%BCAPI%E3%81%AE%E3%82%B5%E3%83%B3%E3%83%97%E3%83%AB%E3%82%B3%E3%83%BC%E3%83%89
  65         script_tag = re.search(r'<script\s*[^>]+?src="https://ssl-cache\.stream\.ne\.jp/(?P<host>www\d+)/(?P<publisher>[a-z0-9]+)/[^"]+?/if\.js"', webpage)
  66         if not script_tag:
  67             return
  68         host, publisher = script_tag.groups()
  69         for m in re.finditer(r'(?s)PlayerFactoryIF\.create\(\s*({[^\}]+?})\s*\)\s*;', webpage):
  70             # TODO: using json.loads here as InfoExtractor._parse_json is not classmethod
  71             info = json.loads(js_to_json(m.group(1)))
  72             mid = base64.b64decode(info.get('m')).decode()
  73             yield f'jstream:{host}:{publisher}:{mid}'