yt_dlp/extractor/amp.py

   1 from .common import InfoExtractor
   2 from ..utils import (
   3     determine_ext,
   4     ExtractorError,
   5     int_or_none,
   6     mimetype2ext,
   7     parse_iso8601,
   8     unified_timestamp,
   9     url_or_none,
  10 )
  11
  12
  13 class AMPIE(InfoExtractor):  # XXX: Conventionally, base classes should end with BaseIE/InfoExtractor
  14     # parse Akamai Adaptive Media Player feed
  15     def _extract_feed_info(self, url):
  16         feed = self._download_json(
  17             url, None, 'Downloading Akamai AMP feed',
  18             'Unable to download Akamai AMP feed')
  19         item = feed.get('channel', {}).get('item')
  20         if not item:
  21             raise ExtractorError('%s said: %s' % (self.IE_NAME, feed['error']))
  22
  23         video_id = item['guid']
  24
  25         def get_media_node(name, default=None):
  26             media_name = 'media-%s' % name
  27             media_group = item.get('media-group') or item
  28             return media_group.get(media_name) or item.get(media_name) or item.get(name, default)
  29
  30         thumbnails = []
  31         media_thumbnail = get_media_node('thumbnail')
  32         if media_thumbnail:
  33             if isinstance(media_thumbnail, dict):
  34                 media_thumbnail = [media_thumbnail]
  35             for thumbnail_data in media_thumbnail:
  36                 thumbnail = thumbnail_data.get('@attributes', {})
  37                 thumbnail_url = url_or_none(thumbnail.get('url'))
  38                 if not thumbnail_url:
  39                     continue
  40                 thumbnails.append({
  41                     'url': self._proto_relative_url(thumbnail_url, 'http:'),
  42                     'width': int_or_none(thumbnail.get('width')),
  43                     'height': int_or_none(thumbnail.get('height')),
  44                 })
  45
  46         subtitles = {}
  47         media_subtitle = get_media_node('subTitle')
  48         if media_subtitle:
  49             if isinstance(media_subtitle, dict):
  50                 media_subtitle = [media_subtitle]
  51             for subtitle_data in media_subtitle:
  52                 subtitle = subtitle_data.get('@attributes', {})
  53                 subtitle_href = url_or_none(subtitle.get('href'))
  54                 if not subtitle_href:
  55                     continue
  56                 subtitles.setdefault(subtitle.get('lang') or 'en', []).append({
  57                     'url': subtitle_href,
  58                     'ext': mimetype2ext(subtitle.get('type')) or determine_ext(subtitle_href),
  59                 })
  60
  61         formats = []
  62         media_content = get_media_node('content')
  63         if isinstance(media_content, dict):
  64             media_content = [media_content]
  65         for media_data in media_content:
  66             media = media_data.get('@attributes', {})
  67             media_url = url_or_none(media.get('url'))
  68             if not media_url:
  69                 continue
  70             ext = mimetype2ext(media.get('type')) or determine_ext(media_url)
  71             if ext == 'f4m':
  72                 formats.extend(self._extract_f4m_formats(
  73                     media_url + '?hdcore=3.4.0&plugin=aasp-3.4.0.132.124',
  74                     video_id, f4m_id='hds', fatal=False))
  75             elif ext == 'm3u8':
  76                 formats.extend(self._extract_m3u8_formats(
  77                     media_url, video_id, 'mp4', m3u8_id='hls', fatal=False))
  78             else:
  79                 formats.append({
  80                     'format_id': media_data.get('media-category', {}).get('@attributes', {}).get('label'),
  81                     'url': media_url,
  82                     'tbr': int_or_none(media.get('bitrate')),
  83                     'filesize': int_or_none(media.get('fileSize')),
  84                     'ext': ext,
  85                 })
  86
  87         timestamp = unified_timestamp(item.get('pubDate'), ' ') or parse_iso8601(item.get('dc-date'))
  88
  89         return {
  90             'id': video_id,
  91             'title': get_media_node('title'),
  92             'description': get_media_node('description'),
  93             'thumbnails': thumbnails,
  94             'timestamp': timestamp,
  95             'duration': int_or_none(media_content[0].get('@attributes', {}).get('duration')),
  96             'subtitles': subtitles,
  97             'formats': formats,
  98         }