yt_dlp/extractor/amp.py

   1 from .common import InfoExtractor
   2 from ..utils import (
   3     ExtractorError,
   4     determine_ext,
   5     int_or_none,
   6     mimetype2ext,
   7     parse_iso8601,
   8     strip_jsonp,
   9     unified_timestamp,
  10     url_or_none,
  11 )
  12
  13
  14 class AMPIE(InfoExtractor):  # XXX: Conventionally, base classes should end with BaseIE/InfoExtractor
  15     # parse Akamai Adaptive Media Player feed
  16     def _extract_feed_info(self, url):
  17         feed = self._download_json(
  18             url, None, 'Downloading Akamai AMP feed',
  19             'Unable to download Akamai AMP feed', transform_source=strip_jsonp)
  20         item = feed.get('channel', {}).get('item')
  21         if not item:
  22             raise ExtractorError('{} said: {}'.format(self.IE_NAME, feed['error']))
  23
  24         video_id = item['guid']
  25
  26         def get_media_node(name, default=None):
  27             media_name = f'media-{name}'
  28             media_group = item.get('media-group') or item
  29             return media_group.get(media_name) or item.get(media_name) or item.get(name, default)
  30
  31         thumbnails = []
  32         media_thumbnail = get_media_node('thumbnail')
  33         if media_thumbnail:
  34             if isinstance(media_thumbnail, dict):
  35                 media_thumbnail = [media_thumbnail]
  36             for thumbnail_data in media_thumbnail:
  37                 thumbnail = thumbnail_data.get('@attributes', {})
  38                 thumbnail_url = url_or_none(thumbnail.get('url'))
  39                 if not thumbnail_url:
  40                     continue
  41                 thumbnails.append({
  42                     'url': self._proto_relative_url(thumbnail_url, 'http:'),
  43                     'width': int_or_none(thumbnail.get('width')),
  44                     'height': int_or_none(thumbnail.get('height')),
  45                 })
  46
  47         subtitles = {}
  48         media_subtitle = get_media_node('subTitle')
  49         if media_subtitle:
  50             if isinstance(media_subtitle, dict):
  51                 media_subtitle = [media_subtitle]
  52             for subtitle_data in media_subtitle:
  53                 subtitle = subtitle_data.get('@attributes', {})
  54                 subtitle_href = url_or_none(subtitle.get('href'))
  55                 if not subtitle_href:
  56                     continue
  57                 subtitles.setdefault(subtitle.get('lang') or 'en', []).append({
  58                     'url': subtitle_href,
  59                     'ext': mimetype2ext(subtitle.get('type')) or determine_ext(subtitle_href),
  60                 })
  61
  62         formats = []
  63         media_content = get_media_node('content')
  64         if isinstance(media_content, dict):
  65             media_content = [media_content]
  66         for media_data in media_content:
  67             media = media_data.get('@attributes', {})
  68             media_url = url_or_none(media.get('url'))
  69             if not media_url:
  70                 continue
  71             ext = mimetype2ext(media.get('type')) or determine_ext(media_url)
  72             if ext == 'f4m':
  73                 formats.extend(self._extract_f4m_formats(
  74                     media_url + '?hdcore=3.4.0&plugin=aasp-3.4.0.132.124',
  75                     video_id, f4m_id='hds', fatal=False))
  76             elif ext == 'm3u8':
  77                 fmts, subs = self._extract_m3u8_formats_and_subtitles(
  78                     media_url, video_id, 'mp4', m3u8_id='hls', fatal=False)
  79                 formats.extend(fmts)
  80                 self._merge_subtitles(subs, target=subtitles)
  81             else:
  82                 formats.append({
  83                     'format_id': media_data.get('media-category', {}).get('@attributes', {}).get('label'),
  84                     'url': media_url,
  85                     'tbr': int_or_none(media.get('bitrate')),
  86                     'filesize': int_or_none(media.get('fileSize')),
  87                     'ext': ext,
  88                 })
  89
  90         timestamp = unified_timestamp(item.get('pubDate'), ' ') or parse_iso8601(item.get('dc-date'))
  91
  92         return {
  93             'id': video_id,
  94             'title': get_media_node('title'),
  95             'description': get_media_node('description'),
  96             'thumbnails': thumbnails,
  97             'timestamp': timestamp,
  98             'duration': int_or_none(media_content[0].get('@attributes', {}).get('duration')),
  99             'subtitles': subtitles,
 100             'formats': formats,
 101         }